Cleveland Clinic Journal of Medicine

The United States preventive services task force (USPSTF) recently threw cold water on the use of calcium and vitamin D supplements to prevent fractures in adults, either finding inadequate evidence to make a recommendation or recommending against supplementation, depending on the population and the doses used.¹

Complicating this issue, several recent studies have raised concern about the long-term cardiovascular risk of calcium supplementation.

With so many people taking calcium supplements, how do we put this into context for our patients? I believe that we need to consider the whole person when discussing these supplements, as there are data that they also help reduce the risk of falls, cancer, and even overall mortality rates.

THE USPSTF’S METHODS

The USPSTF bases its recommendations on explicit criteria² developed by its Evidence-based Practice Center, which is under contract to the US Agency for Healthcare Research and Quality to conduct systematic reviews of the evidence on specific topics in clinical prevention. The USPSTF grades the strength of the evidence for the effectiveness of specific clinical preventive services as:

• A (strongly recommended)
• B (recommended)
• C (no recommendation)
• D (recommended against)
• I (insufficient evidence to make a recommendation for or against).

USPSTF recommendations consider the evidence of both benefit and harm of the intervention but do not include the cost of the intervention in the assessment.³

THE USPSTF’S GRADES ON CALCIUM AND VITAMIN D SUPPLEMENTATION

The USPSTF made the following recommendations in February 2013 about the use of calcium and vitamin D supplementation:

• For primary prevention of fractures in premenopausal women and men: grade I (current evidence is insufficient to assess the balance of the benefits and harms)
• For primary prevention of fractures in noninstitutionalized postmenopausal women, in daily doses greater than 400 IU of vitamin D and greater than 1,000 mg of calcium: also grade I
• For primary prevention of fractures in noninstitutionalized postmenopausal women, in daily doses of 400 IU or less of vitamin D and 1,000 mg or less of calcium: grade D (the USPSTF recommends against it, as these doses increase the incidence of renal stones and there is “adequate” evidence that these doses have no effect on the incidence of fractures).

WHAT THE USPSTF DID NOT DISCUSS

These recommendations do not apply to everybody. Rather, the document discusses “the effectiveness of specific clinical preventive services for patients without related signs or symptoms,”¹ and states that the recommendations do not pertain to patients with osteoporosis or vitamin D deficiency or those who have had fractures.

Also, the document does not discuss the use of calcium supplementation by itself in fracture prevention. nor does it discuss possible benefits of calcium and vitamin D other than fracture prevention, such as reducing the risk of falls, cancer, or death. Further, the document states that “appropriate intake” of vitamin D and calcium is “essential to overall health”¹ but does not state the amount that is considered appropriate.

The document does refer the reader to other USPSTF recommendations concerning screening for osteoporosis in women age 65 and older and in younger women who demonstrate the fracture risk of a 65-year-old woman,⁴ as well as to its recommendation for vitamin D supplementation to prevent falls in community-dwelling adults age 65 and older who are at higher risk of falls.⁵

Not included: A new meta-analysis

The USPSTF document also notes that after their review was completed, another metaanalysis concluded that fracture risk may be reduced by taking vitamin D in doses of 800 IU or higher.⁶

In that study, Bischoff-Ferrari et al⁶ performed a pooled analysis of vitamin D dose requirements for fracture prevention from 11 double-blind, randomized, controlled trials of oral vitamin D supplementation taken either daily or at weekly or 4-month intervals with or without calcium, compared with placebo or calcium alone in people age 65 and older. The primary end points were the incidence of hip fracture and any nonvertebral fracture according to Cox regression analysis, with adjustment for age, sex, community or institutional dwelling, and study. The aim was to evaluate actual vitamin D intake rather than the assigned dosage groups in the trials.

On the basis of actual vitamin D intake, the incidence of hip fracture was significantly (30%) lower in people with the highest actual intake (792–2,000 IU per day) than in controls. There was no reduction in the risk of hip fracture at any actual intake levels lower than 792 IU per day. Using this same analytic technique, the reduction in the incidence of nonvertebral fracture at the highest actual intake level was 16%.

Why were their findings different than those of the USPSTF? The authors hypothesized that some previous high-quality trials of vitamin D supplementation either showed no benefit because the participants were noncompliant and thus took less than the intended dose of vitamin D, or showed an unexpected benefit because the participants actually took more vitamin D than was specified in the study.

The USPSTF recommendations did not include studies of vitamin D without calcium, whereas Bischoff-Ferrari et al did, which could also explain some of the differences in the findings, as not all of the studies included in the two documents were the same. Several previous meta-analyses suggested that the dose of vitamin D was irrelevant when vitamin D was combined with calcium.

The data from Bischoff-Ferrari et al suggested that at the highest actual intake level of vitamin D, a smaller amount of calcium supplementation (< 1,000 mg daily) may be more beneficial in reducing fracture risk than a larger amount (≥ 1,000 mg daily). This is important, given the current level of concern initially raised by Bolland et al⁷ and others about the possible risks of higher doses of calcium supplements increasing cardiovascular risk. (More on this below.)

WHAT OTHER ORGANIZATIONS SAY

Both the National Osteoporosis Foundation and the American Society of Bone and Mineral Research suggest following the 2010 recommendations of the Institute of Medicine⁸ on calcium and vitamin D instead of those of the USPSTF, as the former address the overall health benefits of calcium and vitamin D in healthy individuals rather than only fracture prevention.

Neither the Institute of Medicine nor the USPSTF, however, addresses vitamin D requirements of people at high risk, such as those with vitamin D deficiency due to very little sun exposure, dark skin, problems absorbing dietary fat, or medications that interfere with vitamin D absorption, or those with osteoporosis.

The Institute of Medicine suggests that, for healthy adults under age 71, an adequate vitamin D intake is 600 IU daily, and for healthy adults age 71 and older it is 800 IU daily. They state that the safe upper limit for daily intake of vitamin D is 4,000 IU. As for adequate calcium intake, the daily recommendation is 1,200 mg for women ages 50 through 70, and 1,200 mg for all adults age 71 and older. As I have already discussed, the Institute of Medicine recommendations are based on the overall health benefits of calcium and vitamin D rather than solely on fracture prevention. Monitoring of vitamin D levels is not recommended unless the patient has osteoporosis or is at risk for vitamin D deficiency.

Risks of calcium supplementation

Much has been written recently about the risks of calcium supplementation.

This concern was first raised in 2008 by Bolland et al⁷ in a post hoc analysis of data collected to evaluate the effect of calcium supplements on bone density and fracture.⁷ More myocardial infarctions occurred in the calcium supplement group than in the placebo group, but the difference was not statistically significant, and the events occurred only in those who took more than 1,000 mg of calcium daily.

The same group reanalyzed data from the Women’s Health Initiative and found a 24% higher risk of myocardial infarction in women who took calcium with or without vitamin D, but only in those women assigned to take calcium supplementation who had not taken calcium supplements before the study began.⁹

More recently, Xiao et al¹⁰ evaluated the effect of both dietary and supplemental calcium on cardiovascular disease mortality rates.¹⁰ This was a prospective study of 388,229 men and women who participated in the National Institutes of Health-American Association of Retired Persons Diet and Heart Study. Supplemental calcium intake was associated with an elevated risk of cardiovascular disease in men, but not in women. Dietary calcium intake was unrelated to cardiovascular death.

The latest study to address this issue was from the Swedish Mammography Cohort, a population-based cohort that included 61,433 women born between 1914 and 1948, with a mean follow-up of 19 years.¹¹ Diet was evaluated by food frequency questionnaires. A daily dietary intake of calcium below 600 mg was associated with higher risks of all-cause mortality, cardiovascular disease, ischemic heart disease, and stroke. However, compared with women whose daily calcium intake was between 600 and 999 mg, a dietary intake of more than 1,400 mg/day was associated with a higher death rate, with a hazard ratio for all-cause mortality of 1.40, cardiovascular disease 1.49, and ischemic heart disease 2.14.

Unfortunately, none of these studies were designed to assess the risk of cardiovascular disease related to calcium supplementation. Like the USPSTF, both the National Osteoporosis Foundation and the American Society of Bone and Mineral Research state that this type of study is needed to clarify both the benefit and risk of calcium supplementation.

Until these data are available, the American Society of Bone and Mineral Research has advised doctors and their patients “to discuss the best strategy for each individual patient, putting supplements as the last resort for healthier adults if they cannot reach recommended levels through the intake of calcium and vitamin rich foods.” For adults who cannot tolerate dairy products, calcium can be obtained from calcium-supplemented foods such as orange juice and Jello and from nondairy sources such as leafy green vegetables, almonds, garbanzo beans, tofu, and eggs.¹²

The National Osteoporosis Foundation suggests following the Institute of Medicine recommendations for adequate calcium and vitamin D rather than the USPSTF recommendations, most likely because the former are based on the overall health benefits of calcium and vitamin D rather than fracture prevention only. However, it reminds us that the Institute of Medicine recommendations do not apply to patients who are at the highest risk of fracture, ie, those with osteoporosis and vitamin D deficiency.

TAKE-HOME POINTS

• All medications, including those available over the counter, have benefits and risks.
• Even the USPSTF states that for a healthy lifestyle, the diet should contain adequate calcium and vitamin D intake.
• When following guidelines, practitioners should be certain that the guidelines pertain to the population they are treating—for example, not to apply the Institute of Medicine recommendations to a woman with a hip fracture, but that a healthy premenopausal woman who is taking calcium supplements should be advised to stop the supplements and focus on dietary sources of calcium.
• Only if individuals cannot obtain the recommended amount of calcium in their diet is it advisable for them to take a calcium supplement.

My recommendations

Based on the information summarized above, I recommend that my patients obtain as much calcium as possible from their diet—between 600 and 1,200 mg daily—and to take a calcium supplement only if they cannot obtain that amount of calcium in the diet. However, 24-hour calcium excretion is not recommended as a marker of calcium intake.

I also advise my patients to take a vitamin D supplement, per the Institute of Medicine report for overall good health. The USPSTF recommendations concerning vitamin D and calcium address only fracture prevention. As I am responsible for the overall health of my patients, not just fracture prevention, I choose to follow the National Osteoporosis Foundation and Institute of Medicine recommendations, not those of the USPSTF.

The United States preventive services task force (USPSTF) recently threw cold water on the use of calcium and vitamin D supplements to prevent fractures in adults, either finding inadequate evidence to make a recommendation or recommending against supplementation, depending on the population and the doses used.¹

Complicating this issue, several recent studies have raised concern about the long-term cardiovascular risk of calcium supplementation.

With so many people taking calcium supplements, how do we put this into context for our patients? I believe that we need to consider the whole person when discussing these supplements, as there are data that they also help reduce the risk of falls, cancer, and even overall mortality rates.

THE USPSTF’S METHODS

The USPSTF bases its recommendations on explicit criteria² developed by its Evidence-based Practice Center, which is under contract to the US Agency for Healthcare Research and Quality to conduct systematic reviews of the evidence on specific topics in clinical prevention. The USPSTF grades the strength of the evidence for the effectiveness of specific clinical preventive services as:

• A (strongly recommended)
• B (recommended)
• C (no recommendation)
• D (recommended against)
• I (insufficient evidence to make a recommendation for or against).

USPSTF recommendations consider the evidence of both benefit and harm of the intervention but do not include the cost of the intervention in the assessment.³

THE USPSTF’S GRADES ON CALCIUM AND VITAMIN D SUPPLEMENTATION

The USPSTF made the following recommendations in February 2013 about the use of calcium and vitamin D supplementation:

• For primary prevention of fractures in premenopausal women and men: grade I (current evidence is insufficient to assess the balance of the benefits and harms)
• For primary prevention of fractures in noninstitutionalized postmenopausal women, in daily doses greater than 400 IU of vitamin D and greater than 1,000 mg of calcium: also grade I
• For primary prevention of fractures in noninstitutionalized postmenopausal women, in daily doses of 400 IU or less of vitamin D and 1,000 mg or less of calcium: grade D (the USPSTF recommends against it, as these doses increase the incidence of renal stones and there is “adequate” evidence that these doses have no effect on the incidence of fractures).

WHAT THE USPSTF DID NOT DISCUSS

These recommendations do not apply to everybody. Rather, the document discusses “the effectiveness of specific clinical preventive services for patients without related signs or symptoms,”¹ and states that the recommendations do not pertain to patients with osteoporosis or vitamin D deficiency or those who have had fractures.

Also, the document does not discuss the use of calcium supplementation by itself in fracture prevention. nor does it discuss possible benefits of calcium and vitamin D other than fracture prevention, such as reducing the risk of falls, cancer, or death. Further, the document states that “appropriate intake” of vitamin D and calcium is “essential to overall health”¹ but does not state the amount that is considered appropriate.

The document does refer the reader to other USPSTF recommendations concerning screening for osteoporosis in women age 65 and older and in younger women who demonstrate the fracture risk of a 65-year-old woman,⁴ as well as to its recommendation for vitamin D supplementation to prevent falls in community-dwelling adults age 65 and older who are at higher risk of falls.⁵

Not included: A new meta-analysis

The USPSTF document also notes that after their review was completed, another metaanalysis concluded that fracture risk may be reduced by taking vitamin D in doses of 800 IU or higher.⁶

In that study, Bischoff-Ferrari et al⁶ performed a pooled analysis of vitamin D dose requirements for fracture prevention from 11 double-blind, randomized, controlled trials of oral vitamin D supplementation taken either daily or at weekly or 4-month intervals with or without calcium, compared with placebo or calcium alone in people age 65 and older. The primary end points were the incidence of hip fracture and any nonvertebral fracture according to Cox regression analysis, with adjustment for age, sex, community or institutional dwelling, and study. The aim was to evaluate actual vitamin D intake rather than the assigned dosage groups in the trials.

On the basis of actual vitamin D intake, the incidence of hip fracture was significantly (30%) lower in people with the highest actual intake (792–2,000 IU per day) than in controls. There was no reduction in the risk of hip fracture at any actual intake levels lower than 792 IU per day. Using this same analytic technique, the reduction in the incidence of nonvertebral fracture at the highest actual intake level was 16%.

Why were their findings different than those of the USPSTF? The authors hypothesized that some previous high-quality trials of vitamin D supplementation either showed no benefit because the participants were noncompliant and thus took less than the intended dose of vitamin D, or showed an unexpected benefit because the participants actually took more vitamin D than was specified in the study.

The USPSTF recommendations did not include studies of vitamin D without calcium, whereas Bischoff-Ferrari et al did, which could also explain some of the differences in the findings, as not all of the studies included in the two documents were the same. Several previous meta-analyses suggested that the dose of vitamin D was irrelevant when vitamin D was combined with calcium.

The data from Bischoff-Ferrari et al suggested that at the highest actual intake level of vitamin D, a smaller amount of calcium supplementation (< 1,000 mg daily) may be more beneficial in reducing fracture risk than a larger amount (≥ 1,000 mg daily). This is important, given the current level of concern initially raised by Bolland et al⁷ and others about the possible risks of higher doses of calcium supplements increasing cardiovascular risk. (More on this below.)

WHAT OTHER ORGANIZATIONS SAY

Both the National Osteoporosis Foundation and the American Society of Bone and Mineral Research suggest following the 2010 recommendations of the Institute of Medicine⁸ on calcium and vitamin D instead of those of the USPSTF, as the former address the overall health benefits of calcium and vitamin D in healthy individuals rather than only fracture prevention.

Neither the Institute of Medicine nor the USPSTF, however, addresses vitamin D requirements of people at high risk, such as those with vitamin D deficiency due to very little sun exposure, dark skin, problems absorbing dietary fat, or medications that interfere with vitamin D absorption, or those with osteoporosis.

The Institute of Medicine suggests that, for healthy adults under age 71, an adequate vitamin D intake is 600 IU daily, and for healthy adults age 71 and older it is 800 IU daily. They state that the safe upper limit for daily intake of vitamin D is 4,000 IU. As for adequate calcium intake, the daily recommendation is 1,200 mg for women ages 50 through 70, and 1,200 mg for all adults age 71 and older. As I have already discussed, the Institute of Medicine recommendations are based on the overall health benefits of calcium and vitamin D rather than solely on fracture prevention. Monitoring of vitamin D levels is not recommended unless the patient has osteoporosis or is at risk for vitamin D deficiency.

Risks of calcium supplementation

Much has been written recently about the risks of calcium supplementation.

This concern was first raised in 2008 by Bolland et al⁷ in a post hoc analysis of data collected to evaluate the effect of calcium supplements on bone density and fracture.⁷ More myocardial infarctions occurred in the calcium supplement group than in the placebo group, but the difference was not statistically significant, and the events occurred only in those who took more than 1,000 mg of calcium daily.

The same group reanalyzed data from the Women’s Health Initiative and found a 24% higher risk of myocardial infarction in women who took calcium with or without vitamin D, but only in those women assigned to take calcium supplementation who had not taken calcium supplements before the study began.⁹

More recently, Xiao et al¹⁰ evaluated the effect of both dietary and supplemental calcium on cardiovascular disease mortality rates.¹⁰ This was a prospective study of 388,229 men and women who participated in the National Institutes of Health-American Association of Retired Persons Diet and Heart Study. Supplemental calcium intake was associated with an elevated risk of cardiovascular disease in men, but not in women. Dietary calcium intake was unrelated to cardiovascular death.

The latest study to address this issue was from the Swedish Mammography Cohort, a population-based cohort that included 61,433 women born between 1914 and 1948, with a mean follow-up of 19 years.¹¹ Diet was evaluated by food frequency questionnaires. A daily dietary intake of calcium below 600 mg was associated with higher risks of all-cause mortality, cardiovascular disease, ischemic heart disease, and stroke. However, compared with women whose daily calcium intake was between 600 and 999 mg, a dietary intake of more than 1,400 mg/day was associated with a higher death rate, with a hazard ratio for all-cause mortality of 1.40, cardiovascular disease 1.49, and ischemic heart disease 2.14.

Unfortunately, none of these studies were designed to assess the risk of cardiovascular disease related to calcium supplementation. Like the USPSTF, both the National Osteoporosis Foundation and the American Society of Bone and Mineral Research state that this type of study is needed to clarify both the benefit and risk of calcium supplementation.

Until these data are available, the American Society of Bone and Mineral Research has advised doctors and their patients “to discuss the best strategy for each individual patient, putting supplements as the last resort for healthier adults if they cannot reach recommended levels through the intake of calcium and vitamin rich foods.” For adults who cannot tolerate dairy products, calcium can be obtained from calcium-supplemented foods such as orange juice and Jello and from nondairy sources such as leafy green vegetables, almonds, garbanzo beans, tofu, and eggs.¹²

The National Osteoporosis Foundation suggests following the Institute of Medicine recommendations for adequate calcium and vitamin D rather than the USPSTF recommendations, most likely because the former are based on the overall health benefits of calcium and vitamin D rather than fracture prevention only. However, it reminds us that the Institute of Medicine recommendations do not apply to patients who are at the highest risk of fracture, ie, those with osteoporosis and vitamin D deficiency.

TAKE-HOME POINTS

• All medications, including those available over the counter, have benefits and risks.
• Even the USPSTF states that for a healthy lifestyle, the diet should contain adequate calcium and vitamin D intake.
• When following guidelines, practitioners should be certain that the guidelines pertain to the population they are treating—for example, not to apply the Institute of Medicine recommendations to a woman with a hip fracture, but that a healthy premenopausal woman who is taking calcium supplements should be advised to stop the supplements and focus on dietary sources of calcium.
• Only if individuals cannot obtain the recommended amount of calcium in their diet is it advisable for them to take a calcium supplement.

My recommendations

Based on the information summarized above, I recommend that my patients obtain as much calcium as possible from their diet—between 600 and 1,200 mg daily—and to take a calcium supplement only if they cannot obtain that amount of calcium in the diet. However, 24-hour calcium excretion is not recommended as a marker of calcium intake.

I also advise my patients to take a vitamin D supplement, per the Institute of Medicine report for overall good health. The USPSTF recommendations concerning vitamin D and calcium address only fracture prevention. As I am responsible for the overall health of my patients, not just fracture prevention, I choose to follow the National Osteoporosis Foundation and Institute of Medicine recommendations, not those of the USPSTF.

The United States preventive services task force (USPSTF) recently threw cold water on the use of calcium and vitamin D supplements to prevent fractures in adults, either finding inadequate evidence to make a recommendation or recommending against supplementation, depending on the population and the doses used.¹

Complicating this issue, several recent studies have raised concern about the long-term cardiovascular risk of calcium supplementation.

With so many people taking calcium supplements, how do we put this into context for our patients? I believe that we need to consider the whole person when discussing these supplements, as there are data that they also help reduce the risk of falls, cancer, and even overall mortality rates.

THE USPSTF’S METHODS

The USPSTF bases its recommendations on explicit criteria² developed by its Evidence-based Practice Center, which is under contract to the US Agency for Healthcare Research and Quality to conduct systematic reviews of the evidence on specific topics in clinical prevention. The USPSTF grades the strength of the evidence for the effectiveness of specific clinical preventive services as:

• A (strongly recommended)
• B (recommended)
• C (no recommendation)
• D (recommended against)
• I (insufficient evidence to make a recommendation for or against).

USPSTF recommendations consider the evidence of both benefit and harm of the intervention but do not include the cost of the intervention in the assessment.³

THE USPSTF’S GRADES ON CALCIUM AND VITAMIN D SUPPLEMENTATION

The USPSTF made the following recommendations in February 2013 about the use of calcium and vitamin D supplementation:

• For primary prevention of fractures in premenopausal women and men: grade I (current evidence is insufficient to assess the balance of the benefits and harms)
• For primary prevention of fractures in noninstitutionalized postmenopausal women, in daily doses greater than 400 IU of vitamin D and greater than 1,000 mg of calcium: also grade I
• For primary prevention of fractures in noninstitutionalized postmenopausal women, in daily doses of 400 IU or less of vitamin D and 1,000 mg or less of calcium: grade D (the USPSTF recommends against it, as these doses increase the incidence of renal stones and there is “adequate” evidence that these doses have no effect on the incidence of fractures).

WHAT THE USPSTF DID NOT DISCUSS

These recommendations do not apply to everybody. Rather, the document discusses “the effectiveness of specific clinical preventive services for patients without related signs or symptoms,”¹ and states that the recommendations do not pertain to patients with osteoporosis or vitamin D deficiency or those who have had fractures.

Also, the document does not discuss the use of calcium supplementation by itself in fracture prevention. nor does it discuss possible benefits of calcium and vitamin D other than fracture prevention, such as reducing the risk of falls, cancer, or death. Further, the document states that “appropriate intake” of vitamin D and calcium is “essential to overall health”¹ but does not state the amount that is considered appropriate.

The document does refer the reader to other USPSTF recommendations concerning screening for osteoporosis in women age 65 and older and in younger women who demonstrate the fracture risk of a 65-year-old woman,⁴ as well as to its recommendation for vitamin D supplementation to prevent falls in community-dwelling adults age 65 and older who are at higher risk of falls.⁵

Not included: A new meta-analysis

The USPSTF document also notes that after their review was completed, another metaanalysis concluded that fracture risk may be reduced by taking vitamin D in doses of 800 IU or higher.⁶

In that study, Bischoff-Ferrari et al⁶ performed a pooled analysis of vitamin D dose requirements for fracture prevention from 11 double-blind, randomized, controlled trials of oral vitamin D supplementation taken either daily or at weekly or 4-month intervals with or without calcium, compared with placebo or calcium alone in people age 65 and older. The primary end points were the incidence of hip fracture and any nonvertebral fracture according to Cox regression analysis, with adjustment for age, sex, community or institutional dwelling, and study. The aim was to evaluate actual vitamin D intake rather than the assigned dosage groups in the trials.

On the basis of actual vitamin D intake, the incidence of hip fracture was significantly (30%) lower in people with the highest actual intake (792–2,000 IU per day) than in controls. There was no reduction in the risk of hip fracture at any actual intake levels lower than 792 IU per day. Using this same analytic technique, the reduction in the incidence of nonvertebral fracture at the highest actual intake level was 16%.

Why were their findings different than those of the USPSTF? The authors hypothesized that some previous high-quality trials of vitamin D supplementation either showed no benefit because the participants were noncompliant and thus took less than the intended dose of vitamin D, or showed an unexpected benefit because the participants actually took more vitamin D than was specified in the study.

The USPSTF recommendations did not include studies of vitamin D without calcium, whereas Bischoff-Ferrari et al did, which could also explain some of the differences in the findings, as not all of the studies included in the two documents were the same. Several previous meta-analyses suggested that the dose of vitamin D was irrelevant when vitamin D was combined with calcium.

The data from Bischoff-Ferrari et al suggested that at the highest actual intake level of vitamin D, a smaller amount of calcium supplementation (< 1,000 mg daily) may be more beneficial in reducing fracture risk than a larger amount (≥ 1,000 mg daily). This is important, given the current level of concern initially raised by Bolland et al⁷ and others about the possible risks of higher doses of calcium supplements increasing cardiovascular risk. (More on this below.)

WHAT OTHER ORGANIZATIONS SAY

Both the National Osteoporosis Foundation and the American Society of Bone and Mineral Research suggest following the 2010 recommendations of the Institute of Medicine⁸ on calcium and vitamin D instead of those of the USPSTF, as the former address the overall health benefits of calcium and vitamin D in healthy individuals rather than only fracture prevention.

Neither the Institute of Medicine nor the USPSTF, however, addresses vitamin D requirements of people at high risk, such as those with vitamin D deficiency due to very little sun exposure, dark skin, problems absorbing dietary fat, or medications that interfere with vitamin D absorption, or those with osteoporosis.

The Institute of Medicine suggests that, for healthy adults under age 71, an adequate vitamin D intake is 600 IU daily, and for healthy adults age 71 and older it is 800 IU daily. They state that the safe upper limit for daily intake of vitamin D is 4,000 IU. As for adequate calcium intake, the daily recommendation is 1,200 mg for women ages 50 through 70, and 1,200 mg for all adults age 71 and older. As I have already discussed, the Institute of Medicine recommendations are based on the overall health benefits of calcium and vitamin D rather than solely on fracture prevention. Monitoring of vitamin D levels is not recommended unless the patient has osteoporosis or is at risk for vitamin D deficiency.

Risks of calcium supplementation

Much has been written recently about the risks of calcium supplementation.

This concern was first raised in 2008 by Bolland et al⁷ in a post hoc analysis of data collected to evaluate the effect of calcium supplements on bone density and fracture.⁷ More myocardial infarctions occurred in the calcium supplement group than in the placebo group, but the difference was not statistically significant, and the events occurred only in those who took more than 1,000 mg of calcium daily.

The same group reanalyzed data from the Women’s Health Initiative and found a 24% higher risk of myocardial infarction in women who took calcium with or without vitamin D, but only in those women assigned to take calcium supplementation who had not taken calcium supplements before the study began.⁹

More recently, Xiao et al¹⁰ evaluated the effect of both dietary and supplemental calcium on cardiovascular disease mortality rates.¹⁰ This was a prospective study of 388,229 men and women who participated in the National Institutes of Health-American Association of Retired Persons Diet and Heart Study. Supplemental calcium intake was associated with an elevated risk of cardiovascular disease in men, but not in women. Dietary calcium intake was unrelated to cardiovascular death.

The latest study to address this issue was from the Swedish Mammography Cohort, a population-based cohort that included 61,433 women born between 1914 and 1948, with a mean follow-up of 19 years.¹¹ Diet was evaluated by food frequency questionnaires. A daily dietary intake of calcium below 600 mg was associated with higher risks of all-cause mortality, cardiovascular disease, ischemic heart disease, and stroke. However, compared with women whose daily calcium intake was between 600 and 999 mg, a dietary intake of more than 1,400 mg/day was associated with a higher death rate, with a hazard ratio for all-cause mortality of 1.40, cardiovascular disease 1.49, and ischemic heart disease 2.14.

Unfortunately, none of these studies were designed to assess the risk of cardiovascular disease related to calcium supplementation. Like the USPSTF, both the National Osteoporosis Foundation and the American Society of Bone and Mineral Research state that this type of study is needed to clarify both the benefit and risk of calcium supplementation.

Until these data are available, the American Society of Bone and Mineral Research has advised doctors and their patients “to discuss the best strategy for each individual patient, putting supplements as the last resort for healthier adults if they cannot reach recommended levels through the intake of calcium and vitamin rich foods.” For adults who cannot tolerate dairy products, calcium can be obtained from calcium-supplemented foods such as orange juice and Jello and from nondairy sources such as leafy green vegetables, almonds, garbanzo beans, tofu, and eggs.¹²

The National Osteoporosis Foundation suggests following the Institute of Medicine recommendations for adequate calcium and vitamin D rather than the USPSTF recommendations, most likely because the former are based on the overall health benefits of calcium and vitamin D rather than fracture prevention only. However, it reminds us that the Institute of Medicine recommendations do not apply to patients who are at the highest risk of fracture, ie, those with osteoporosis and vitamin D deficiency.

TAKE-HOME POINTS

• All medications, including those available over the counter, have benefits and risks.
• Even the USPSTF states that for a healthy lifestyle, the diet should contain adequate calcium and vitamin D intake.
• When following guidelines, practitioners should be certain that the guidelines pertain to the population they are treating—for example, not to apply the Institute of Medicine recommendations to a woman with a hip fracture, but that a healthy premenopausal woman who is taking calcium supplements should be advised to stop the supplements and focus on dietary sources of calcium.
• Only if individuals cannot obtain the recommended amount of calcium in their diet is it advisable for them to take a calcium supplement.

My recommendations

Based on the information summarized above, I recommend that my patients obtain as much calcium as possible from their diet—between 600 and 1,200 mg daily—and to take a calcium supplement only if they cannot obtain that amount of calcium in the diet. However, 24-hour calcium excretion is not recommended as a marker of calcium intake.

I also advise my patients to take a vitamin D supplement, per the Institute of Medicine report for overall good health. The USPSTF recommendations concerning vitamin D and calcium address only fracture prevention. As I am responsible for the overall health of my patients, not just fracture prevention, I choose to follow the National Osteoporosis Foundation and Institute of Medicine recommendations, not those of the USPSTF.

To the Editor: We recently read the article by Dr. Emmanuel Bravo.¹ In his comprehensive paper, he defined a road map for the workup of resistant hypertension. Resistant hypertension is a challenging problem in everyday practice, with multiple pitfalls at each step from diagnosis to treatment.

Although not mentioned in the paper, obstructive sleep apnea is strongly associated with hypertension, and its prevalence in patients with resistant hypertension can be as high as 83%.² The upper airway resistance syndrome is another form of sleep-disordered breathing in which transient increases in upper airway resistance result in repetitive electroencephalographic arousals. Unlike obstructive sleep apnea, upper airway resistance syndrome is not associated with apnea or diminished airflow, although snoring and excessive daytime somnolence are common. Repeated arousals, desaturations, or both during sleep lead to recurrent sympathetic surges with resultant nocturnal hypertension. There are a number of reports in the literature of large blood-pressure reductions after continuous positive airway pressure treatment.³

In conclusion, sleep-disordered breathing syndromes should be sought vigorously in cases of resistant hypertension, and every effort should be taken for proper management.

To the Editor: We recently read the article by Dr. Emmanuel Bravo.¹ In his comprehensive paper, he defined a road map for the workup of resistant hypertension. Resistant hypertension is a challenging problem in everyday practice, with multiple pitfalls at each step from diagnosis to treatment.

Although not mentioned in the paper, obstructive sleep apnea is strongly associated with hypertension, and its prevalence in patients with resistant hypertension can be as high as 83%.² The upper airway resistance syndrome is another form of sleep-disordered breathing in which transient increases in upper airway resistance result in repetitive electroencephalographic arousals. Unlike obstructive sleep apnea, upper airway resistance syndrome is not associated with apnea or diminished airflow, although snoring and excessive daytime somnolence are common. Repeated arousals, desaturations, or both during sleep lead to recurrent sympathetic surges with resultant nocturnal hypertension. There are a number of reports in the literature of large blood-pressure reductions after continuous positive airway pressure treatment.³

In conclusion, sleep-disordered breathing syndromes should be sought vigorously in cases of resistant hypertension, and every effort should be taken for proper management.

To the Editor: We recently read the article by Dr. Emmanuel Bravo.¹ In his comprehensive paper, he defined a road map for the workup of resistant hypertension. Resistant hypertension is a challenging problem in everyday practice, with multiple pitfalls at each step from diagnosis to treatment.

Although not mentioned in the paper, obstructive sleep apnea is strongly associated with hypertension, and its prevalence in patients with resistant hypertension can be as high as 83%.² The upper airway resistance syndrome is another form of sleep-disordered breathing in which transient increases in upper airway resistance result in repetitive electroencephalographic arousals. Unlike obstructive sleep apnea, upper airway resistance syndrome is not associated with apnea or diminished airflow, although snoring and excessive daytime somnolence are common. Repeated arousals, desaturations, or both during sleep lead to recurrent sympathetic surges with resultant nocturnal hypertension. There are a number of reports in the literature of large blood-pressure reductions after continuous positive airway pressure treatment.³

In conclusion, sleep-disordered breathing syndromes should be sought vigorously in cases of resistant hypertension, and every effort should be taken for proper management.

Experts have argued for decades about how best to manage opiate dependence, with practitioners generally subscribing to one of two strategies: either total abstinence or medication-assisted treatment (MAT).

Although MAT has proven efficacy, it has been slow to gain acceptance, and the gold standard of care since the 1930s has been abstinence-based treatment. Among elite institutional holdouts against MAT was the Hazelden Treatment Center, a leading treatment institution and publishing house that had been wedded to the abstinence model since it was founded in 1949.¹ Now, Hazelden has gone on record as embracing MAT, raising the possibility that the two predominant treatment philosophies for opiate-dependent patients may no longer be at odds.

FROM ABSTINENCE TO METHADONE MAINTENANCE

The modern day abstinence-based movement in this country started in the decade before the founding of Hazelden. In 1935, the US government opened the first of two federal drug treatment centers, known as the United States Narcotic Farm, in Lexington, KY.² The move by the government to get into the addiction treatment business largely stemmed from frustration over the growing problem of addiction at that time, coupled with a dearth of treatment options for addicts in the wake of the 1914 Harrison Narcotics Act.

The Narcotic Farm was an impressive facility—for all intents and purposes, a specialized prison—that initially housed 1,200 people. In addition to prisoners, it also accepted voluntary, nonprisoner patients. In many ways, it was ahead of its time. It offered a wide variety of services, including detoxification, group therapy, individual therapy, psychiatric and medical services, and vocational rehabilitation.² Housed on the premises was the Addiction Research Center at Lexington, the first intramural research branch of the National Institute of Mental Health. After the “Blue Grass” mandatory commitment laws were passed in the 1940s, even the voluntary patients were ultimately committed for a 1-year sentence at Lexington. This facility, and its sister facility in Ft. Worth, TX, would have been the envy of any modern-day abstinence-based treatment center in terms of the services offered and the long lengths of stay.

The quality of the program, as evidenced by the impressive array of services and long stays, would lead one to expect that its treatment outcomes over nearly 40 years of operation were equally stellar. However, in terms of outcomes the Farm was an abysmal failure, as shown by numerous studies demonstrating relapse rates of more than 90% in the patients discharged from it.^2,3

Similar frustrations at other abstinence-based treatment centers from the 1940s through the 1960s led Dr. Vincent Dole, the “father of methadone maintenance,” to conclude in 1971 that after detoxification from opiates, “human addicts almost always return to use of narcotics after they leave the hospital where they have been detoxified.”⁴ That realization inspired Dr. Dole and his wife and colleague Dr. Marie Nyswander to revisit the idea of medication-assisted treatment, an approach previously used by the morphine maintenance clinics of the early 1900s. This work led to the development of government-sanctioned methadone clinics across America and to the realization that long-term recovery was possible with medication, even without a lengthy hospital stay. For this revolutionary work on opiate addiction, Dr. Dole won the prestigious Lasker Award in 1988.

The major reason for the success of methadone was that, because of its pharmacokinetic profile, it could stabilize the patient through once-daily dosing without sedation or narcosis. As noted by Dr. Dole, once patients are on a stable dosing regimen, the obsessive preoccupation with drug use fades away.⁵

Despite its success, methadone maintenance had its share of detractors. It was fraught with controversy because it was viewed as a crutch, and those who were on it were often not considered by their abstinent peers as being in true recovery. The reasons for the negative attitudes toward MAT are unclear but may reflect antiquated beliefs that addiction may be indicative of a failure of morals or will, and that patients ought to be able to simply stop using.

Whatever the reason for the animosity surrounding MAT, it should be noted that an expert consensus panel convened by the Betty Ford Center in 2007 agreed that patients on MAT met their consensus definition of sobriety.⁶ The issue of what constitutes recovery remains a very complex and hotly debated topic that is beyond the scope of this paper and that has been discussed elsewhere.^6,7

For more than 3 decades, methadone was the only medication available for MAT. Federal regulations limit the dispensing of methadone to licensed clinics, most of which are located in major metropolitan areas. Patients must go to the clinic every day to receive their dose of methadone—a major inconvenience, especially to those with transportation issues. Adding to the lack of appeal of methadone maintenance is that the clinics are typically located in the higher-crime areas of cities. Savvy drug dealers know the location of these clinics and often loiter on nearby street corners in an attempt to lure addicts away from recovery by flaunting their illicit drugs.

A final, very significant drawback of methadone is its safety profile. It is a full-agonist narcotic that can be fatal in overdose or in the induction phase, especially if taken with other drugs, such as benzodiazepines.

2003: BUPRENORPHINE-NALOXONE IS APPROVED

Such concerns led researchers to search for other medications to be used for MAT that could perhaps be prescribed in a typical outpatient physician practice. For many reasons, buprenorphine became the most promising candidate. In 2003, the US Food and Drug Administration approved the combination medication buprenorphine-naloxone (Sub-oxone) as only the second drug indicated for maintenance treatment of opioid dependence in the United States.

Buprenorphine differs from methadone in that it is a partial agonist at mu opiate receptors, and therefore has a “ceiling” or “plateau” effect in terms of dose-response and a much improved safety profile. Unlike methadone, buprenorphine can be prescribed in a doctor’s office and does not have to be dispensed at a government-approved clinic.

Unfortunately, buprenorphine-maintained patients seem to carry the same stigma in the recovery community as those maintained on methadone—that they are simply substituting one drug for another. Detractors usually fail to consider that, as with methadone, patients do not report getting “high” from taking buprenorphine. Patients will often state that when they first start taking it, they “feel something,” but after a few days of adjustment, they simply feel normal. They don’t feel high, they are no longer in withdrawal, their cravings are virtually eliminated, and their opiate receptors are effectively occupied and blocked, so there is no “high” in the event of a relapse.

What’s more, buprenorphine is not a medication that will help them deal with life’s stressors by “chemical coping.” Sober coping is a skill they must learn by actively participating in a solid 12-step-based recovery program and, in some cases, in psychotherapy. By removing the drug obsession, buprenorphine promotes and facilitates the important recovery goal of learning how to deal with life on life’s terms.

ADDICTION AS CHRONIC ILLNESS

Outcomes studies of addiction treatment have focused largely on rates of relapse after discharge from acute treatments such as residential rehabilitation, partial hospitalization, and intensive outpatient programs. With MAT, however, outcomes research has primarily looked at the duration of retention in treatment.

The change in focus between the two types of treatment coincides with a paradigm shift that views addiction as a chronic condition that requires ongoing care. Continued participation in prescribed care with demonstrated efficacy is considered to be the major indicator of success. Under the chronic illness model employed by MAT providers, if a patient reverted to briefly using a drug of abuse, this would be an issue to address in his ongoing treatment and would not necessarily indicate treatment failure as with the acute care model. Beyond retention rates, research has demonstrated that MAT with methadone results in reductions in rates of criminal activity, illicit drug use, acquisition of human immunodeficiency virus, and overall mortality.^8–10

In outcomes studies, MAT has repeatedly shown better efficacy than abstinence-based approaches. During the first 5 years of its implementation, in 4,000 patients, methadone maintenance boasted 1-year retention rates exceeding 98%.¹¹ Over the subsequent 3 years, with the number of patients approaching 35,000, the 1-year retention rates fell to around 60%—still far exceeding results of abstinence-based treatment and approximating the number cited in most modern studies.¹¹

The retention rates in buprenorphine programs are similarly promising. Studies of 12 to 13 weeks duration have shown retention rates of 52% to 79%.^12–15 Six-month studies have demonstrated retention rates of 43% to 100%.^16–19 Another study showed that 38% of opiate-dependent patients remained in treatment with buprenorphine at 5 years.²⁰ Surprisingly, most of the buprenorphine studies have been conducted in office-based practices, which are less structured than outpatient methadone programs.

MEDICATION-ASSISTED TREATMENT IS GAINING ACCEPTANCE

Data from decades of experience with MAT strongly support the conclusion that it is superior to abstinence-based approaches.

The importance of a patient staying in treatment cannot be overemphasized, as the consequence of failing in recovery may well be an early death. On average, heroin addicts lose about 18 years of life expectancy, and the mortality rate for injection users is roughly 2% per year.²¹ The mortality rate for heroin users is 6 to 20 times greater than for age-matched peers who are not drug users.²²

As high as these numbers are, they are even higher for abusers of prescription narcotics. The annual death rate associated with opioid pain relievers (4.8 per 100,000) is nearly double that associated with illicit drugs (2.8 per 100,000).²³

The recent and rather radical change in treatment philosophy by Hazelden came as a shock to some, a disappointment to others, and a welcome change to many who saw this as a move by one of the more respected treatment centers in the country to fall in line with the body of evidence that supports MAT for those suffering from opiate dependence. It remains a mystery why so many, if not most, addiction treatment centers in the United States cling to the abstinence-based philosophy despite the overwhelming data from decades of research and experience that show that abstinence does not work for the majority of opiate addicts.

Complete abstinence from opiate drugs of abuse and potentially addictive medications is a noble but perhaps unreachable goal for many sufferers. Hazelden’s announced acceptance of MAT gives credence to the value of recovery goals that are not entirely drug-free.

Dr. Dole was correct in stating that opiate addicts usually return to drugs if not provided with MAT. Treatment programs need to inform opiate-dependent patients that abstinence-based treatment offers only a 1 in 10 chance of success. Perhaps some patients, armed with the daunting statistics regarding abstinence, will be inspired to devote themselves wholeheartedly to their recovery in an effort to make it into that elite 10% group that achieves long-lasting recovery without the aid of medications. But for the other 90%, it is encouraging to hear that Hazelden, the model treatment center for most abstinence-based programs in this country, may now lead other abstinence-based centers to reconsider their treatment philosophies.

Historically, US doctors were not allowed by federal law to prescribe opiates for addiction treatment. With the passage of DATA 2000, buprenorphine (alone or in combination with naloxone) can be prescribed for addiction treatment only by providers who obtain a waiver from the US Drug Enforcement Administration (DEA). Any doctor can become qualified to prescribe buprenorphine or buprenorphinenaloxone after completing an 8-hour online training course (available at www.buppractice.com and at www.aaap.org/buprenorphine) and by obtaining a DATA 2000 waiver and a new prescribing number from the DEA. Doctors are initially limited to treating only 30 patients with buprenorphine-naloxone at any given time, but can apply for an extension to 100 patients after having had their waiver for 1 year.

As MAT continues to gain favor, demand will grow for more providers to obtain their waivers to prescribe buprenorphine and buprenorphine-naloxone. Historically, there have always been too few methadone clinics to meet the demand. One can hope that the growing number of waivered providers will greatly improve access to care by opiate addicts, no matter where they reside. Qualified prescribers of buprenorphine and buprenorphine-naloxone are limited by the federal restrictions on the numbers of patients they can treat. If the chronic disease of addiction is to be integrated into the continuing-care approach of modern medicine and managed alongside other chronic diseases, primary care providers who are not specialized in treating addiction will need to be become comfortable with maintaining patients on buprenorphine-naloxone.⁷ Presumably, such patients will have already been stabilized through participation in addiction treatment programs in their respective geographic areas. Primary care providers will need to develop relationships with local addictionologists and treatment programs so that they will be able to refer those in active addiction for induction and stabilization on MAT and will be able to refer those already stabilized on MAT back to such specialists when relapses occur.

We may finally be approaching a time when structured residential treatment and MAT are not mutually exclusive options for our patients. These treatment options must work together for optimal outcomes. Based on our experience with hundreds of patients at Cleveland Clinic’s Alcohol and Drug Recovery Center, we believe this change of treatment philosophy is long overdue. In clinical settings, patients do not fit cleanly into one treatment arm or another and often require a blended approach to effect long-lasting change. Hazelden’s shift of treatment philosophy is an indication that this research-supported viewpoint is gaining acceptance in the traditionally drug-free halls of addiction treatment programs.

Experts have argued for decades about how best to manage opiate dependence, with practitioners generally subscribing to one of two strategies: either total abstinence or medication-assisted treatment (MAT).

Although MAT has proven efficacy, it has been slow to gain acceptance, and the gold standard of care since the 1930s has been abstinence-based treatment. Among elite institutional holdouts against MAT was the Hazelden Treatment Center, a leading treatment institution and publishing house that had been wedded to the abstinence model since it was founded in 1949.¹ Now, Hazelden has gone on record as embracing MAT, raising the possibility that the two predominant treatment philosophies for opiate-dependent patients may no longer be at odds.

FROM ABSTINENCE TO METHADONE MAINTENANCE

The modern day abstinence-based movement in this country started in the decade before the founding of Hazelden. In 1935, the US government opened the first of two federal drug treatment centers, known as the United States Narcotic Farm, in Lexington, KY.² The move by the government to get into the addiction treatment business largely stemmed from frustration over the growing problem of addiction at that time, coupled with a dearth of treatment options for addicts in the wake of the 1914 Harrison Narcotics Act.

The Narcotic Farm was an impressive facility—for all intents and purposes, a specialized prison—that initially housed 1,200 people. In addition to prisoners, it also accepted voluntary, nonprisoner patients. In many ways, it was ahead of its time. It offered a wide variety of services, including detoxification, group therapy, individual therapy, psychiatric and medical services, and vocational rehabilitation.² Housed on the premises was the Addiction Research Center at Lexington, the first intramural research branch of the National Institute of Mental Health. After the “Blue Grass” mandatory commitment laws were passed in the 1940s, even the voluntary patients were ultimately committed for a 1-year sentence at Lexington. This facility, and its sister facility in Ft. Worth, TX, would have been the envy of any modern-day abstinence-based treatment center in terms of the services offered and the long lengths of stay.

The quality of the program, as evidenced by the impressive array of services and long stays, would lead one to expect that its treatment outcomes over nearly 40 years of operation were equally stellar. However, in terms of outcomes the Farm was an abysmal failure, as shown by numerous studies demonstrating relapse rates of more than 90% in the patients discharged from it.^2,3

Similar frustrations at other abstinence-based treatment centers from the 1940s through the 1960s led Dr. Vincent Dole, the “father of methadone maintenance,” to conclude in 1971 that after detoxification from opiates, “human addicts almost always return to use of narcotics after they leave the hospital where they have been detoxified.”⁴ That realization inspired Dr. Dole and his wife and colleague Dr. Marie Nyswander to revisit the idea of medication-assisted treatment, an approach previously used by the morphine maintenance clinics of the early 1900s. This work led to the development of government-sanctioned methadone clinics across America and to the realization that long-term recovery was possible with medication, even without a lengthy hospital stay. For this revolutionary work on opiate addiction, Dr. Dole won the prestigious Lasker Award in 1988.

The major reason for the success of methadone was that, because of its pharmacokinetic profile, it could stabilize the patient through once-daily dosing without sedation or narcosis. As noted by Dr. Dole, once patients are on a stable dosing regimen, the obsessive preoccupation with drug use fades away.⁵

Despite its success, methadone maintenance had its share of detractors. It was fraught with controversy because it was viewed as a crutch, and those who were on it were often not considered by their abstinent peers as being in true recovery. The reasons for the negative attitudes toward MAT are unclear but may reflect antiquated beliefs that addiction may be indicative of a failure of morals or will, and that patients ought to be able to simply stop using.

Whatever the reason for the animosity surrounding MAT, it should be noted that an expert consensus panel convened by the Betty Ford Center in 2007 agreed that patients on MAT met their consensus definition of sobriety.⁶ The issue of what constitutes recovery remains a very complex and hotly debated topic that is beyond the scope of this paper and that has been discussed elsewhere.^6,7

For more than 3 decades, methadone was the only medication available for MAT. Federal regulations limit the dispensing of methadone to licensed clinics, most of which are located in major metropolitan areas. Patients must go to the clinic every day to receive their dose of methadone—a major inconvenience, especially to those with transportation issues. Adding to the lack of appeal of methadone maintenance is that the clinics are typically located in the higher-crime areas of cities. Savvy drug dealers know the location of these clinics and often loiter on nearby street corners in an attempt to lure addicts away from recovery by flaunting their illicit drugs.

A final, very significant drawback of methadone is its safety profile. It is a full-agonist narcotic that can be fatal in overdose or in the induction phase, especially if taken with other drugs, such as benzodiazepines.

2003: BUPRENORPHINE-NALOXONE IS APPROVED

Such concerns led researchers to search for other medications to be used for MAT that could perhaps be prescribed in a typical outpatient physician practice. For many reasons, buprenorphine became the most promising candidate. In 2003, the US Food and Drug Administration approved the combination medication buprenorphine-naloxone (Sub-oxone) as only the second drug indicated for maintenance treatment of opioid dependence in the United States.

Buprenorphine differs from methadone in that it is a partial agonist at mu opiate receptors, and therefore has a “ceiling” or “plateau” effect in terms of dose-response and a much improved safety profile. Unlike methadone, buprenorphine can be prescribed in a doctor’s office and does not have to be dispensed at a government-approved clinic.

Unfortunately, buprenorphine-maintained patients seem to carry the same stigma in the recovery community as those maintained on methadone—that they are simply substituting one drug for another. Detractors usually fail to consider that, as with methadone, patients do not report getting “high” from taking buprenorphine. Patients will often state that when they first start taking it, they “feel something,” but after a few days of adjustment, they simply feel normal. They don’t feel high, they are no longer in withdrawal, their cravings are virtually eliminated, and their opiate receptors are effectively occupied and blocked, so there is no “high” in the event of a relapse.

What’s more, buprenorphine is not a medication that will help them deal with life’s stressors by “chemical coping.” Sober coping is a skill they must learn by actively participating in a solid 12-step-based recovery program and, in some cases, in psychotherapy. By removing the drug obsession, buprenorphine promotes and facilitates the important recovery goal of learning how to deal with life on life’s terms.

ADDICTION AS CHRONIC ILLNESS

Outcomes studies of addiction treatment have focused largely on rates of relapse after discharge from acute treatments such as residential rehabilitation, partial hospitalization, and intensive outpatient programs. With MAT, however, outcomes research has primarily looked at the duration of retention in treatment.

The change in focus between the two types of treatment coincides with a paradigm shift that views addiction as a chronic condition that requires ongoing care. Continued participation in prescribed care with demonstrated efficacy is considered to be the major indicator of success. Under the chronic illness model employed by MAT providers, if a patient reverted to briefly using a drug of abuse, this would be an issue to address in his ongoing treatment and would not necessarily indicate treatment failure as with the acute care model. Beyond retention rates, research has demonstrated that MAT with methadone results in reductions in rates of criminal activity, illicit drug use, acquisition of human immunodeficiency virus, and overall mortality.^8–10

In outcomes studies, MAT has repeatedly shown better efficacy than abstinence-based approaches. During the first 5 years of its implementation, in 4,000 patients, methadone maintenance boasted 1-year retention rates exceeding 98%.¹¹ Over the subsequent 3 years, with the number of patients approaching 35,000, the 1-year retention rates fell to around 60%—still far exceeding results of abstinence-based treatment and approximating the number cited in most modern studies.¹¹

The retention rates in buprenorphine programs are similarly promising. Studies of 12 to 13 weeks duration have shown retention rates of 52% to 79%.^12–15 Six-month studies have demonstrated retention rates of 43% to 100%.^16–19 Another study showed that 38% of opiate-dependent patients remained in treatment with buprenorphine at 5 years.²⁰ Surprisingly, most of the buprenorphine studies have been conducted in office-based practices, which are less structured than outpatient methadone programs.

MEDICATION-ASSISTED TREATMENT IS GAINING ACCEPTANCE

Data from decades of experience with MAT strongly support the conclusion that it is superior to abstinence-based approaches.

The importance of a patient staying in treatment cannot be overemphasized, as the consequence of failing in recovery may well be an early death. On average, heroin addicts lose about 18 years of life expectancy, and the mortality rate for injection users is roughly 2% per year.²¹ The mortality rate for heroin users is 6 to 20 times greater than for age-matched peers who are not drug users.²²

As high as these numbers are, they are even higher for abusers of prescription narcotics. The annual death rate associated with opioid pain relievers (4.8 per 100,000) is nearly double that associated with illicit drugs (2.8 per 100,000).²³

The recent and rather radical change in treatment philosophy by Hazelden came as a shock to some, a disappointment to others, and a welcome change to many who saw this as a move by one of the more respected treatment centers in the country to fall in line with the body of evidence that supports MAT for those suffering from opiate dependence. It remains a mystery why so many, if not most, addiction treatment centers in the United States cling to the abstinence-based philosophy despite the overwhelming data from decades of research and experience that show that abstinence does not work for the majority of opiate addicts.

Complete abstinence from opiate drugs of abuse and potentially addictive medications is a noble but perhaps unreachable goal for many sufferers. Hazelden’s announced acceptance of MAT gives credence to the value of recovery goals that are not entirely drug-free.

Dr. Dole was correct in stating that opiate addicts usually return to drugs if not provided with MAT. Treatment programs need to inform opiate-dependent patients that abstinence-based treatment offers only a 1 in 10 chance of success. Perhaps some patients, armed with the daunting statistics regarding abstinence, will be inspired to devote themselves wholeheartedly to their recovery in an effort to make it into that elite 10% group that achieves long-lasting recovery without the aid of medications. But for the other 90%, it is encouraging to hear that Hazelden, the model treatment center for most abstinence-based programs in this country, may now lead other abstinence-based centers to reconsider their treatment philosophies.

Historically, US doctors were not allowed by federal law to prescribe opiates for addiction treatment. With the passage of DATA 2000, buprenorphine (alone or in combination with naloxone) can be prescribed for addiction treatment only by providers who obtain a waiver from the US Drug Enforcement Administration (DEA). Any doctor can become qualified to prescribe buprenorphine or buprenorphinenaloxone after completing an 8-hour online training course (available at www.buppractice.com and at www.aaap.org/buprenorphine) and by obtaining a DATA 2000 waiver and a new prescribing number from the DEA. Doctors are initially limited to treating only 30 patients with buprenorphine-naloxone at any given time, but can apply for an extension to 100 patients after having had their waiver for 1 year.

As MAT continues to gain favor, demand will grow for more providers to obtain their waivers to prescribe buprenorphine and buprenorphine-naloxone. Historically, there have always been too few methadone clinics to meet the demand. One can hope that the growing number of waivered providers will greatly improve access to care by opiate addicts, no matter where they reside. Qualified prescribers of buprenorphine and buprenorphine-naloxone are limited by the federal restrictions on the numbers of patients they can treat. If the chronic disease of addiction is to be integrated into the continuing-care approach of modern medicine and managed alongside other chronic diseases, primary care providers who are not specialized in treating addiction will need to be become comfortable with maintaining patients on buprenorphine-naloxone.⁷ Presumably, such patients will have already been stabilized through participation in addiction treatment programs in their respective geographic areas. Primary care providers will need to develop relationships with local addictionologists and treatment programs so that they will be able to refer those in active addiction for induction and stabilization on MAT and will be able to refer those already stabilized on MAT back to such specialists when relapses occur.

We may finally be approaching a time when structured residential treatment and MAT are not mutually exclusive options for our patients. These treatment options must work together for optimal outcomes. Based on our experience with hundreds of patients at Cleveland Clinic’s Alcohol and Drug Recovery Center, we believe this change of treatment philosophy is long overdue. In clinical settings, patients do not fit cleanly into one treatment arm or another and often require a blended approach to effect long-lasting change. Hazelden’s shift of treatment philosophy is an indication that this research-supported viewpoint is gaining acceptance in the traditionally drug-free halls of addiction treatment programs.

Experts have argued for decades about how best to manage opiate dependence, with practitioners generally subscribing to one of two strategies: either total abstinence or medication-assisted treatment (MAT).

Although MAT has proven efficacy, it has been slow to gain acceptance, and the gold standard of care since the 1930s has been abstinence-based treatment. Among elite institutional holdouts against MAT was the Hazelden Treatment Center, a leading treatment institution and publishing house that had been wedded to the abstinence model since it was founded in 1949.¹ Now, Hazelden has gone on record as embracing MAT, raising the possibility that the two predominant treatment philosophies for opiate-dependent patients may no longer be at odds.

FROM ABSTINENCE TO METHADONE MAINTENANCE

The modern day abstinence-based movement in this country started in the decade before the founding of Hazelden. In 1935, the US government opened the first of two federal drug treatment centers, known as the United States Narcotic Farm, in Lexington, KY.² The move by the government to get into the addiction treatment business largely stemmed from frustration over the growing problem of addiction at that time, coupled with a dearth of treatment options for addicts in the wake of the 1914 Harrison Narcotics Act.

The Narcotic Farm was an impressive facility—for all intents and purposes, a specialized prison—that initially housed 1,200 people. In addition to prisoners, it also accepted voluntary, nonprisoner patients. In many ways, it was ahead of its time. It offered a wide variety of services, including detoxification, group therapy, individual therapy, psychiatric and medical services, and vocational rehabilitation.² Housed on the premises was the Addiction Research Center at Lexington, the first intramural research branch of the National Institute of Mental Health. After the “Blue Grass” mandatory commitment laws were passed in the 1940s, even the voluntary patients were ultimately committed for a 1-year sentence at Lexington. This facility, and its sister facility in Ft. Worth, TX, would have been the envy of any modern-day abstinence-based treatment center in terms of the services offered and the long lengths of stay.

The quality of the program, as evidenced by the impressive array of services and long stays, would lead one to expect that its treatment outcomes over nearly 40 years of operation were equally stellar. However, in terms of outcomes the Farm was an abysmal failure, as shown by numerous studies demonstrating relapse rates of more than 90% in the patients discharged from it.^2,3

Similar frustrations at other abstinence-based treatment centers from the 1940s through the 1960s led Dr. Vincent Dole, the “father of methadone maintenance,” to conclude in 1971 that after detoxification from opiates, “human addicts almost always return to use of narcotics after they leave the hospital where they have been detoxified.”⁴ That realization inspired Dr. Dole and his wife and colleague Dr. Marie Nyswander to revisit the idea of medication-assisted treatment, an approach previously used by the morphine maintenance clinics of the early 1900s. This work led to the development of government-sanctioned methadone clinics across America and to the realization that long-term recovery was possible with medication, even without a lengthy hospital stay. For this revolutionary work on opiate addiction, Dr. Dole won the prestigious Lasker Award in 1988.

The major reason for the success of methadone was that, because of its pharmacokinetic profile, it could stabilize the patient through once-daily dosing without sedation or narcosis. As noted by Dr. Dole, once patients are on a stable dosing regimen, the obsessive preoccupation with drug use fades away.⁵

Despite its success, methadone maintenance had its share of detractors. It was fraught with controversy because it was viewed as a crutch, and those who were on it were often not considered by their abstinent peers as being in true recovery. The reasons for the negative attitudes toward MAT are unclear but may reflect antiquated beliefs that addiction may be indicative of a failure of morals or will, and that patients ought to be able to simply stop using.

Whatever the reason for the animosity surrounding MAT, it should be noted that an expert consensus panel convened by the Betty Ford Center in 2007 agreed that patients on MAT met their consensus definition of sobriety.⁶ The issue of what constitutes recovery remains a very complex and hotly debated topic that is beyond the scope of this paper and that has been discussed elsewhere.^6,7

For more than 3 decades, methadone was the only medication available for MAT. Federal regulations limit the dispensing of methadone to licensed clinics, most of which are located in major metropolitan areas. Patients must go to the clinic every day to receive their dose of methadone—a major inconvenience, especially to those with transportation issues. Adding to the lack of appeal of methadone maintenance is that the clinics are typically located in the higher-crime areas of cities. Savvy drug dealers know the location of these clinics and often loiter on nearby street corners in an attempt to lure addicts away from recovery by flaunting their illicit drugs.

A final, very significant drawback of methadone is its safety profile. It is a full-agonist narcotic that can be fatal in overdose or in the induction phase, especially if taken with other drugs, such as benzodiazepines.

2003: BUPRENORPHINE-NALOXONE IS APPROVED

Such concerns led researchers to search for other medications to be used for MAT that could perhaps be prescribed in a typical outpatient physician practice. For many reasons, buprenorphine became the most promising candidate. In 2003, the US Food and Drug Administration approved the combination medication buprenorphine-naloxone (Sub-oxone) as only the second drug indicated for maintenance treatment of opioid dependence in the United States.

Buprenorphine differs from methadone in that it is a partial agonist at mu opiate receptors, and therefore has a “ceiling” or “plateau” effect in terms of dose-response and a much improved safety profile. Unlike methadone, buprenorphine can be prescribed in a doctor’s office and does not have to be dispensed at a government-approved clinic.

Unfortunately, buprenorphine-maintained patients seem to carry the same stigma in the recovery community as those maintained on methadone—that they are simply substituting one drug for another. Detractors usually fail to consider that, as with methadone, patients do not report getting “high” from taking buprenorphine. Patients will often state that when they first start taking it, they “feel something,” but after a few days of adjustment, they simply feel normal. They don’t feel high, they are no longer in withdrawal, their cravings are virtually eliminated, and their opiate receptors are effectively occupied and blocked, so there is no “high” in the event of a relapse.

What’s more, buprenorphine is not a medication that will help them deal with life’s stressors by “chemical coping.” Sober coping is a skill they must learn by actively participating in a solid 12-step-based recovery program and, in some cases, in psychotherapy. By removing the drug obsession, buprenorphine promotes and facilitates the important recovery goal of learning how to deal with life on life’s terms.

ADDICTION AS CHRONIC ILLNESS

Outcomes studies of addiction treatment have focused largely on rates of relapse after discharge from acute treatments such as residential rehabilitation, partial hospitalization, and intensive outpatient programs. With MAT, however, outcomes research has primarily looked at the duration of retention in treatment.

The change in focus between the two types of treatment coincides with a paradigm shift that views addiction as a chronic condition that requires ongoing care. Continued participation in prescribed care with demonstrated efficacy is considered to be the major indicator of success. Under the chronic illness model employed by MAT providers, if a patient reverted to briefly using a drug of abuse, this would be an issue to address in his ongoing treatment and would not necessarily indicate treatment failure as with the acute care model. Beyond retention rates, research has demonstrated that MAT with methadone results in reductions in rates of criminal activity, illicit drug use, acquisition of human immunodeficiency virus, and overall mortality.^8–10

In outcomes studies, MAT has repeatedly shown better efficacy than abstinence-based approaches. During the first 5 years of its implementation, in 4,000 patients, methadone maintenance boasted 1-year retention rates exceeding 98%.¹¹ Over the subsequent 3 years, with the number of patients approaching 35,000, the 1-year retention rates fell to around 60%—still far exceeding results of abstinence-based treatment and approximating the number cited in most modern studies.¹¹

The retention rates in buprenorphine programs are similarly promising. Studies of 12 to 13 weeks duration have shown retention rates of 52% to 79%.^12–15 Six-month studies have demonstrated retention rates of 43% to 100%.^16–19 Another study showed that 38% of opiate-dependent patients remained in treatment with buprenorphine at 5 years.²⁰ Surprisingly, most of the buprenorphine studies have been conducted in office-based practices, which are less structured than outpatient methadone programs.

MEDICATION-ASSISTED TREATMENT IS GAINING ACCEPTANCE

Data from decades of experience with MAT strongly support the conclusion that it is superior to abstinence-based approaches.

The importance of a patient staying in treatment cannot be overemphasized, as the consequence of failing in recovery may well be an early death. On average, heroin addicts lose about 18 years of life expectancy, and the mortality rate for injection users is roughly 2% per year.²¹ The mortality rate for heroin users is 6 to 20 times greater than for age-matched peers who are not drug users.²²

As high as these numbers are, they are even higher for abusers of prescription narcotics. The annual death rate associated with opioid pain relievers (4.8 per 100,000) is nearly double that associated with illicit drugs (2.8 per 100,000).²³

The recent and rather radical change in treatment philosophy by Hazelden came as a shock to some, a disappointment to others, and a welcome change to many who saw this as a move by one of the more respected treatment centers in the country to fall in line with the body of evidence that supports MAT for those suffering from opiate dependence. It remains a mystery why so many, if not most, addiction treatment centers in the United States cling to the abstinence-based philosophy despite the overwhelming data from decades of research and experience that show that abstinence does not work for the majority of opiate addicts.

Complete abstinence from opiate drugs of abuse and potentially addictive medications is a noble but perhaps unreachable goal for many sufferers. Hazelden’s announced acceptance of MAT gives credence to the value of recovery goals that are not entirely drug-free.

Dr. Dole was correct in stating that opiate addicts usually return to drugs if not provided with MAT. Treatment programs need to inform opiate-dependent patients that abstinence-based treatment offers only a 1 in 10 chance of success. Perhaps some patients, armed with the daunting statistics regarding abstinence, will be inspired to devote themselves wholeheartedly to their recovery in an effort to make it into that elite 10% group that achieves long-lasting recovery without the aid of medications. But for the other 90%, it is encouraging to hear that Hazelden, the model treatment center for most abstinence-based programs in this country, may now lead other abstinence-based centers to reconsider their treatment philosophies.

Historically, US doctors were not allowed by federal law to prescribe opiates for addiction treatment. With the passage of DATA 2000, buprenorphine (alone or in combination with naloxone) can be prescribed for addiction treatment only by providers who obtain a waiver from the US Drug Enforcement Administration (DEA). Any doctor can become qualified to prescribe buprenorphine or buprenorphinenaloxone after completing an 8-hour online training course (available at www.buppractice.com and at www.aaap.org/buprenorphine) and by obtaining a DATA 2000 waiver and a new prescribing number from the DEA. Doctors are initially limited to treating only 30 patients with buprenorphine-naloxone at any given time, but can apply for an extension to 100 patients after having had their waiver for 1 year.

As MAT continues to gain favor, demand will grow for more providers to obtain their waivers to prescribe buprenorphine and buprenorphine-naloxone. Historically, there have always been too few methadone clinics to meet the demand. One can hope that the growing number of waivered providers will greatly improve access to care by opiate addicts, no matter where they reside. Qualified prescribers of buprenorphine and buprenorphine-naloxone are limited by the federal restrictions on the numbers of patients they can treat. If the chronic disease of addiction is to be integrated into the continuing-care approach of modern medicine and managed alongside other chronic diseases, primary care providers who are not specialized in treating addiction will need to be become comfortable with maintaining patients on buprenorphine-naloxone.⁷ Presumably, such patients will have already been stabilized through participation in addiction treatment programs in their respective geographic areas. Primary care providers will need to develop relationships with local addictionologists and treatment programs so that they will be able to refer those in active addiction for induction and stabilization on MAT and will be able to refer those already stabilized on MAT back to such specialists when relapses occur.

We may finally be approaching a time when structured residential treatment and MAT are not mutually exclusive options for our patients. These treatment options must work together for optimal outcomes. Based on our experience with hundreds of patients at Cleveland Clinic’s Alcohol and Drug Recovery Center, we believe this change of treatment philosophy is long overdue. In clinical settings, patients do not fit cleanly into one treatment arm or another and often require a blended approach to effect long-lasting change. Hazelden’s shift of treatment philosophy is an indication that this research-supported viewpoint is gaining acceptance in the traditionally drug-free halls of addiction treatment programs.

A 57-year-old woman with no history of cardiovascular disease comes to the clinic for her annual evaluation. She does not have diabetes mellitus, but she does have hypertension and chronic osteoarthritis, currently treated with acetaminophen. Additionally, she admits to active tobacco use. Her systolic blood pressure is 130 mm Hg on therapy with hydrochlorothiazide. Her electrocardiogram demonstrates left ventricular hypertrophy. Her low-density lipoprotein (LDL) cholesterol level is 140 mg/dL, and her high-density lipoprotein (HDL) cholesterol level is 50 mg/dL. Should this patient be started on aspirin therapy?

Acetylsalicylic acid (aspirin) is an analgesic, antipyretic, and anti-inflammatory agent, but its more prominent use today is as an antithrombotic agent to treat or prevent cardiovascular events. Its antithrombotic properties are due to its effects on the enzyme cyclooxygenase. However, cyclooxygenase is also involved in regulation of the gastric mucosa, and so aspirin increases the risk of gastrointestinal bleeding.

Approximately 50 million people take aspirin on a daily basis to treat or prevent cardiovascular disease.¹ Of these, at least half are taking more than 100 mg per day,² reflecting the general belief that, for aspirin dosage, “more is better”—which is not true.

Additionally, recommendations about the use of aspirin were based on studies that included relatively few members of several important subgroups, such as people with diabetes without known cardiovascular disease, women, and the elderly, and thus may not reflect appropriate indications and dosages for these groups.

Here, we examine the literature, outline an individualized approach to aspirin therapy, and highlight areas for future study.

HISTORY OF ASPIRIN USE IN CARDIOVASCULAR DISEASE

• 1700s—Willow bark is used as an analgesic.
• 1897—Synthetic aspirin is developed as an antipyretic and anti-inflammatory agent.
• 1974—First landmark trial of aspirin for secondary prevention of myocardial infarction.³
• 1982—Nobel Prize awarded for discovery of aspirin mechanism.
• 1985—US Food and Drug Administration approves aspirin for the treatment and secondary prevention of acute myocardial infarction.
• 1998—The Second International Study of Infarct Survival (ISIS-2) finds that giving aspirin to patients with myocardial infarction within 24 hours of presentation leads to a significant reduction in vascular deaths.⁴

Ongoing uncertainties

Aspirin now carries a class I indication for all patients with suspected myocardial infarction. Since there are an estimated 600,000 new coronary events and 325,000 recurrent ischemic events per year in the United States,⁵ the need for aspirin will continue to remain great. It is also approved to prevent and treat stroke and in patients with unstable angina.

However, questions continue to emerge about aspirin’s dosing and appropriate use in specific populations. The initial prevention trials used a wide range of doses and, as mentioned, included few women, few people with diabetes, and few elderly people. The uncertainties are especially pertinent for patients without known vascular disease, in whom the absolute risk reduction is much less, making the assessment of bleeding risk particularly important. Furthermore, the absolute risk-to-benefit assessment may be different in certain populations.

Guidelines on the use of aspirin to prevent cardiovascular disease (Table 1)^6–10 have evolved to take into account these possible disparities, and studies are taking place to further investigate aspirin use in these groups.

ASPIRIN AND GASTROINTESTINAL BLEEDING

Aspirin’s association with bleeding, particularly gastrointestinal bleeding, was recognized early as a use-limiting side effect. With or without aspirin, gastrointestinal bleeding is a common cause of morbidity and death, with an incidence of approximately 100 per 100,000 bleeding episodes in adults per year for upper gastrointestinal bleeding and 20 to 30 per 100,000 per year for lower gastrointestinal bleeding.^11,12

The standard dosage (ie, 325 mg/day) is associated with a significantly higher risk of gastrointestinal bleeding (including fatal bleeds) than is 75 mg.¹³ However, even with lower doses, the risk of gastrointestinal bleeding is estimated to be twice as high as with no aspirin.¹⁴

And here is the irony: studies have shown that higher doses of aspirin offer no advantage in preventing thrombotic events compared with lower doses.¹⁵ For example, the Clopidogrel Optimal Loading Dose Usage to Reduce Recurrent Events/Organization to Assess Strategies for Ischemic Stroke Syndromes study reported a higher rate of gastrointestinal bleeding with standard-dose aspirin therapy than with low-dose aspirin, with no additional cardiovascular benefit with the higher dose.¹⁶

Furthermore, several other risk factors increase the risk of gastrointestinal bleeding with aspirin use (Table 2). These risk factors are common in the general population but were not necessarily represented in participants in clinical trials. Thus, estimates of risk based on trial data most likely underestimate actual risk in the general population, and therefore, the individual patient’s risk of gastrointestinal bleeding, based on these and other factors, needs to be taken into consideration.

ASPIRIN IN PATIENTS WITH CORONARY ARTERY DISEASE

Randomized clinical trials have validated the benefits of aspirin in secondary prevention of cardiovascular events in patients who have had a myocardial infarction. Patients with coronary disease who withdraw from aspirin therapy or otherwise do not adhere to it have a risk of cardiovascular events three times higher than those who stay with it.¹⁷

Despite the strong data, however, several issues and questions remain about the use of aspirin for secondary prevention.

Bleeding risk must be considered, since gastrointestinal bleeding is associated with a higher risk of death and myocardial infarction in patients with cardiovascular disease.¹⁸ Many patients with coronary disease are on more than one antiplatelet or anticoagulant therapy for concomitant conditions such as atrial fibrillation or because they underwent a percutaneous intervention, which further increases the risk of bleeding.

This bleeding risk is reflected in changes in the most recent recommendations for aspirin dosing after percutaneous coronary intervention. Earlier guidelines advocated use of either 162 or 325 mg after the procedure. However, the most recent update (in 2011) now supports 81 mg for maintenance dosing after intervention.⁷

Patients with coronary disease but without prior myocardial infarction or intervention. Current guidelines recommend 75 to 162 mg of aspirin in all patients with coronary artery disease.⁶ However, this group is diverse and includes patients who have undergone percutaneous coronary intervention, patients with chronic stable angina, and patients with asymptomatic coronary artery disease found on imaging studies. The magnitude of benefit is not clear for those who have no symptoms or who have stable angina.

Most of the evidence supporting aspirin use in chronic angina came from a single trial in Sweden, in which 2,000 patients with chronic stable angina were given either 75 mg daily or placebo. Those who received aspirin had a 34% lower rate of myocardial infarction and sudden death.¹⁹

A substudy of the Physicians’ Health Study, with fewer patients, also noted a significant reduction in the rate of first myocardial infarction. The dose of aspirin in this study was 325 mg every other day.²⁰

In the Women’s Health Initiative Observational Study, 70% of women with stable cardiovascular disease taking aspirin were taking 325 mg daily.²¹ This study demonstrated a significant reduction in the cardiovascular mortality rate, which supports current guidelines, and found no difference in outcomes with doses of 81 mg compared with 325 mg.²¹ This again corroborates that low-dose aspirin is preferential to standard-dose aspirin in women with cardiovascular disease.

These findings have not been validated in larger prospective trials. Thus, current guidelines for aspirin use may reflect extrapolation of aspirin benefit from higher-risk patients to lower-risk patients.

Nevertheless, although the debate continues, it has generally been accepted that in patients who are at high risk of vascular disease or who have had a myocardial infarction, the benefits of aspirin—a 20% relative reduction in vascular events²²—clearly outweigh the risks.

ASPIRIN FOR PRIMARY PREVENTION

Assessing risk vs benefit is more complex when considering populations without known cardiovascular disease.

Only a few studies have specifically evaluated the use of aspirin for primary prevention (Table 3).^23–31 The initial trials were in male physicians in the United Kingdom and the United States in the late 1980s and had somewhat conflicting results. A British study did not find a significant reduction in myocardial infarction,²³ but the US Physician’s Health Study study did: the relative risk was 0.56 (95% confidence interval 0.45–0.70, P < .00001).²⁴ The US study had more than four times the number of participants, used different dosing (325 mg every other day compared with 500 or 300 mg daily in the British study), and had a higher rate of compliance.

Several studies over the next decade demonstrated variable but significant reductions in cardiovascular events as well.^25–27

A meta-analysis of primary prevention trials of aspirin was published in 2009.²² Although the relative risk reduction was similar in primary and secondary prevention, the absolute risk reduction in primary prevention was not nearly as great as in secondary prevention.

These findings are somewhat difficult to interpret, as the component trials included a wide spectrum of patients, ranging from healthy people with no symptoms and no known risk factors to those with limited risk factors. The trials were also performed over several decades during which primary prevention strategies were evolving. Additionally, most of the participants were middle-aged, nondiabetic men, so the results may not necessarily apply to people with diabetes, to women, or to the elderly. Thus, the pooled data in favor of aspirin for primary prevention may not be as broadly applicable to the general population as was once thought.

Aspirin for primary prevention in women

Guidelines for aspirin use in primary prevention were initially thought to be equally applicable to both sexes. However, concerns about the relatively low number of women participating in the studies and the possible mechanistic differences in aspirin efficacy in men vs women prompted further study.

A meta-analysis of randomized controlled trials found that aspirin was associated with a 12% relative reduction in the incidence of cardiovascular events in women and 14% in men. On the other hand, for stroke, the relative risk reduction was 17% in women, while men had no benefit.³²

Most of the women in this meta-analysis were participants in the Women’s Health Study, and they were at low baseline risk.²⁸ Although only about 10% of patients in this study were over age 65, this older group accounted for most of the benefit: these older women had a 26% risk reduction in major adverse cardiovascular events and 30% reduction in stroke.

Thus, for women, aspirin seems to become effective for primary prevention at an older age than in men, and the guidelines have been changed accordingly (Figure 1).

More women should be taking aspirin than actually are. For example, Rivera et al³³ found that only 41% of eligible women were receiving aspirin for primary prevention and 48% of eligible women were receiving it for secondary prevention.

People with diabetes

People with diabetes without overt cardiovascular disease are at higher risk of cardiovascular events than age- and sex-matched controls.³⁴ On the other hand, people with diabetes may be more prone to aspirin resistance and may not derive as much cardiovascular benefit from aspirin.

Early primary prevention studies included few people with diabetes. Subsequent meta-analyses of trials that used a wide range of aspirin doses found a relative risk reduction of 9%, which was not statistically significant.^9,35,36

But there is some evidence that people with diabetes, with³⁷ and without²² coronary disease, may be at higher inherent risk of bleeding than people without diabetes. Although aspirin may not necessarily increase the risk of bleeding in diabetic patients, recent data suggest no benefit in terms of a reduction in vascular events.³⁸

The balance of risk vs benefit for aspirin in this special population is not clear, although some argue that these patients should be treated somewhere on the spectrum of risk between primary and secondary prevention.

The US Preventive Services Task Force did not differentiate between people with or without diabetes in its 2009 guidelines for aspirin for primary prevention.⁸ However, the debate is reflected in a change in 2010 American College of Cardiology/American Diabetes Association guidelines regarding aspirin use in people with diabetes without known cardiovascular disease.³⁹ As opposed to earlier recommendations from these organizations in favor of aspirin for all people with diabetes regardless of 10-year risk, current recommendations advise low-dose aspirin (81–162 mg) for diabetic patients without known vascular disease who have a greater than 10% risk of a cardiovascular event and are not at increased risk of bleeding.

These changes were based on the findings of two trials: the Prevention and Progression of Arterial Disease and Diabetes Trial (POPADAD) and the Japanese Primary Prevention of Atherosclerosis With Aspirin for Diabetes (JPAD) study. These did not show a statistically significant benefit in prevention of cardiovascular events with aspirin.^29,30

After the new guidelines came out, a meta-analysis further bolstered its recommendations. ⁴⁰ In seven randomized clinical trials in 11,000 patients, the relative risk reduction was 9% with aspirin, which did not reach statistical significance.

The use of statins has been increasing, and this trend may have played a role in the marginal benefit of aspirin therapy in these recent studies. In the Japanese trial, approximately 25% of the patients were known to be using a statin; the percentage of statin use was not reported specifically in POPADAD, but both of these studies were published in 2008, when the proportion of diabetic patients taking a statin would be expected to be higher than in earlier primary prevention trials, which were performed primarily in the 1990s. Thus, the beneficial effects of statins may have somewhat diluted the risk reduction attributable to aspirin.

Trials under way in patients with diabetes

The evolving and somewhat conflicting guidelines highlight the need for further study in patients with diabetes. To address this area, two trials are in progress: the Aspirin and Simvastatin Combination for Cardiovascular Events Prevention Trial in Diabetes (ACCEPT-D) and A Study of Cardiovascular Events in Diabetes (ASCEND).^41,42

ACCEPT-D is testing low-dose aspirin (100 mg daily) in diabetic patients who are also on simvastatin. This study also includes prespecified subgroups stratified by sex, age, and baseline lipid levels.

The ASCEND trial will use the same aspirin dose as ACCEPT-D, with a target enrollment of 10,000 patients with diabetes without known vascular disease.

More frequent dosing for people with diabetes?

Although not supported by current guidelines, recent work has suggested that people with diabetes may need more-frequent dosing of aspirin.⁴³ This topic warrants further investigation.

Aspirin as primary prevention in elderly patients

The incidence of cardiovascular events increases with age³⁷—but so does the incidence of gastrointestinal bleeding.⁴⁴ Upper gastrointestinal bleeding is especially worrisome in the elderly, in whom the estimated case-fatality rate is high.¹² Assessment of risk and benefit is particularly important in patients over age 65 without known coronary disease.

Uncertainty about aspirin use in this population is reflected in the most recent US Preventive Services Task Force guidelines, which do not advocate either for or against regular aspirin use for primary prevention in those over the age of 80.

Data on this topic from clinical trials are limited. The Antithrombotic Trialists’ Collaboration (2009) found that although age is associated with a risk of major coronary events similar to that of other traditional risk factors such as diabetes, hypertension, and tobacco use, older age is also associated with the highest risk of major extracranial bleeding.²²

Because of the lack of data in this population, several studies are currently under way. The Aspirin in Reducing Events in the Elderly (ASPREE) trial is studying 100 mg daily in nondiabetic patients without known cardiovascular disease who are age 70 and older.⁴⁵ An additional trial will study patients age 60 to 85 with concurrent diagnoses of hypertension, hyperlipidemia, or diabetes and will test the same aspirin dose as in ASPREE.⁴⁶ These trials should provide further insight into the safety and efficacy of aspirin for primary prevention in the elderly.

FUTURE DIRECTIONS

Aspirin remains a cornerstone of therapy in patients with cardiovascular disease and in secondary prevention of adverse cardiovascular events, but its role in primary prevention remains under scrutiny. Recommendations have evolved to reflect emerging data in special populations, and an algorithm based on Framingham risk assessment in men for myocardial infarction and ischemic stroke assessment in women for assessing appropriateness of aspirin therapy based on currently available guidelines is presented in Figure 1.^6,8,47–49 Targeted studies have advanced our understanding of aspirin use in women, and future studies in people with diabetes and in the elderly should provide further insight into the role of aspirin for primary prevention in these specific groups as well.

Additionally, the range of doses used in clinical studies has propagated the general misperception that higher doses of aspirin are more efficacious. Future studies should continue to use lower doses of aspirin to minimize bleeding risk with an added focus on re-examining its net benefit in the modern era of increasing statin use, which may reduce the absolute risk reduction attributable to aspirin.

One particular area of debate is whether enteric coating can result in functional aspirin resistance. Grosser et al⁵⁰ found that sustained aspirin resistance was rare, and “pseudoresistance” was related to the use of a single enteric-coated aspirin instead of immediate-release aspirin in people who had not been taking aspirin up to then. This complements an earlier study, which found that enteric-coated aspirin had an appropriate effect when given for 7 days.⁵¹ Therefore, for patients who have not been taking aspirin, the first dose should always be immediate-release, not enteric coated.

SHOULD OUR PATIENT RECEIVE ASPIRIN?

The patient we described at the beginning of this article has several risk factors—hypertension, dyslipidemia, left ventricular hypertrophy, and smoking—but no known cardiovascular disease as yet. Her risk of an adverse cardiovascular event appears moderate. However, her 10-year risk of stroke by the Framingham risk calculation is 10%, which would qualify her for aspirin for primary prevention. Of particular note is that the significance of left ventricular hypertrophy as a risk factor for stroke in women is higher than in men and in our case accounts for half of this patient’s risk.

We should explain to the patient that the anticipated benefits of aspirin for stroke prevention outweigh bleeding risks, and thus aspirin therapy would be recommended. However, with her elevated LDL-cholesterol, she may benefit from a statin, which could lessen the relative risk reduction from additional aspirin use.

A 57-year-old woman with no history of cardiovascular disease comes to the clinic for her annual evaluation. She does not have diabetes mellitus, but she does have hypertension and chronic osteoarthritis, currently treated with acetaminophen. Additionally, she admits to active tobacco use. Her systolic blood pressure is 130 mm Hg on therapy with hydrochlorothiazide. Her electrocardiogram demonstrates left ventricular hypertrophy. Her low-density lipoprotein (LDL) cholesterol level is 140 mg/dL, and her high-density lipoprotein (HDL) cholesterol level is 50 mg/dL. Should this patient be started on aspirin therapy?

Acetylsalicylic acid (aspirin) is an analgesic, antipyretic, and anti-inflammatory agent, but its more prominent use today is as an antithrombotic agent to treat or prevent cardiovascular events. Its antithrombotic properties are due to its effects on the enzyme cyclooxygenase. However, cyclooxygenase is also involved in regulation of the gastric mucosa, and so aspirin increases the risk of gastrointestinal bleeding.

Approximately 50 million people take aspirin on a daily basis to treat or prevent cardiovascular disease.¹ Of these, at least half are taking more than 100 mg per day,² reflecting the general belief that, for aspirin dosage, “more is better”—which is not true.

Additionally, recommendations about the use of aspirin were based on studies that included relatively few members of several important subgroups, such as people with diabetes without known cardiovascular disease, women, and the elderly, and thus may not reflect appropriate indications and dosages for these groups.

Here, we examine the literature, outline an individualized approach to aspirin therapy, and highlight areas for future study.

HISTORY OF ASPIRIN USE IN CARDIOVASCULAR DISEASE

• 1700s—Willow bark is used as an analgesic.
• 1897—Synthetic aspirin is developed as an antipyretic and anti-inflammatory agent.
• 1974—First landmark trial of aspirin for secondary prevention of myocardial infarction.³
• 1982—Nobel Prize awarded for discovery of aspirin mechanism.
• 1985—US Food and Drug Administration approves aspirin for the treatment and secondary prevention of acute myocardial infarction.
• 1998—The Second International Study of Infarct Survival (ISIS-2) finds that giving aspirin to patients with myocardial infarction within 24 hours of presentation leads to a significant reduction in vascular deaths.⁴

Ongoing uncertainties

Aspirin now carries a class I indication for all patients with suspected myocardial infarction. Since there are an estimated 600,000 new coronary events and 325,000 recurrent ischemic events per year in the United States,⁵ the need for aspirin will continue to remain great. It is also approved to prevent and treat stroke and in patients with unstable angina.

However, questions continue to emerge about aspirin’s dosing and appropriate use in specific populations. The initial prevention trials used a wide range of doses and, as mentioned, included few women, few people with diabetes, and few elderly people. The uncertainties are especially pertinent for patients without known vascular disease, in whom the absolute risk reduction is much less, making the assessment of bleeding risk particularly important. Furthermore, the absolute risk-to-benefit assessment may be different in certain populations.

Guidelines on the use of aspirin to prevent cardiovascular disease (Table 1)^6–10 have evolved to take into account these possible disparities, and studies are taking place to further investigate aspirin use in these groups.

ASPIRIN AND GASTROINTESTINAL BLEEDING

Aspirin’s association with bleeding, particularly gastrointestinal bleeding, was recognized early as a use-limiting side effect. With or without aspirin, gastrointestinal bleeding is a common cause of morbidity and death, with an incidence of approximately 100 per 100,000 bleeding episodes in adults per year for upper gastrointestinal bleeding and 20 to 30 per 100,000 per year for lower gastrointestinal bleeding.^11,12

The standard dosage (ie, 325 mg/day) is associated with a significantly higher risk of gastrointestinal bleeding (including fatal bleeds) than is 75 mg.¹³ However, even with lower doses, the risk of gastrointestinal bleeding is estimated to be twice as high as with no aspirin.¹⁴

And here is the irony: studies have shown that higher doses of aspirin offer no advantage in preventing thrombotic events compared with lower doses.¹⁵ For example, the Clopidogrel Optimal Loading Dose Usage to Reduce Recurrent Events/Organization to Assess Strategies for Ischemic Stroke Syndromes study reported a higher rate of gastrointestinal bleeding with standard-dose aspirin therapy than with low-dose aspirin, with no additional cardiovascular benefit with the higher dose.¹⁶

Furthermore, several other risk factors increase the risk of gastrointestinal bleeding with aspirin use (Table 2). These risk factors are common in the general population but were not necessarily represented in participants in clinical trials. Thus, estimates of risk based on trial data most likely underestimate actual risk in the general population, and therefore, the individual patient’s risk of gastrointestinal bleeding, based on these and other factors, needs to be taken into consideration.

ASPIRIN IN PATIENTS WITH CORONARY ARTERY DISEASE

Randomized clinical trials have validated the benefits of aspirin in secondary prevention of cardiovascular events in patients who have had a myocardial infarction. Patients with coronary disease who withdraw from aspirin therapy or otherwise do not adhere to it have a risk of cardiovascular events three times higher than those who stay with it.¹⁷

Despite the strong data, however, several issues and questions remain about the use of aspirin for secondary prevention.

Bleeding risk must be considered, since gastrointestinal bleeding is associated with a higher risk of death and myocardial infarction in patients with cardiovascular disease.¹⁸ Many patients with coronary disease are on more than one antiplatelet or anticoagulant therapy for concomitant conditions such as atrial fibrillation or because they underwent a percutaneous intervention, which further increases the risk of bleeding.

This bleeding risk is reflected in changes in the most recent recommendations for aspirin dosing after percutaneous coronary intervention. Earlier guidelines advocated use of either 162 or 325 mg after the procedure. However, the most recent update (in 2011) now supports 81 mg for maintenance dosing after intervention.⁷

Patients with coronary disease but without prior myocardial infarction or intervention. Current guidelines recommend 75 to 162 mg of aspirin in all patients with coronary artery disease.⁶ However, this group is diverse and includes patients who have undergone percutaneous coronary intervention, patients with chronic stable angina, and patients with asymptomatic coronary artery disease found on imaging studies. The magnitude of benefit is not clear for those who have no symptoms or who have stable angina.

Most of the evidence supporting aspirin use in chronic angina came from a single trial in Sweden, in which 2,000 patients with chronic stable angina were given either 75 mg daily or placebo. Those who received aspirin had a 34% lower rate of myocardial infarction and sudden death.¹⁹

A substudy of the Physicians’ Health Study, with fewer patients, also noted a significant reduction in the rate of first myocardial infarction. The dose of aspirin in this study was 325 mg every other day.²⁰

In the Women’s Health Initiative Observational Study, 70% of women with stable cardiovascular disease taking aspirin were taking 325 mg daily.²¹ This study demonstrated a significant reduction in the cardiovascular mortality rate, which supports current guidelines, and found no difference in outcomes with doses of 81 mg compared with 325 mg.²¹ This again corroborates that low-dose aspirin is preferential to standard-dose aspirin in women with cardiovascular disease.

These findings have not been validated in larger prospective trials. Thus, current guidelines for aspirin use may reflect extrapolation of aspirin benefit from higher-risk patients to lower-risk patients.

Nevertheless, although the debate continues, it has generally been accepted that in patients who are at high risk of vascular disease or who have had a myocardial infarction, the benefits of aspirin—a 20% relative reduction in vascular events²²—clearly outweigh the risks.

ASPIRIN FOR PRIMARY PREVENTION

Assessing risk vs benefit is more complex when considering populations without known cardiovascular disease.

Only a few studies have specifically evaluated the use of aspirin for primary prevention (Table 3).^23–31 The initial trials were in male physicians in the United Kingdom and the United States in the late 1980s and had somewhat conflicting results. A British study did not find a significant reduction in myocardial infarction,²³ but the US Physician’s Health Study study did: the relative risk was 0.56 (95% confidence interval 0.45–0.70, P < .00001).²⁴ The US study had more than four times the number of participants, used different dosing (325 mg every other day compared with 500 or 300 mg daily in the British study), and had a higher rate of compliance.

Several studies over the next decade demonstrated variable but significant reductions in cardiovascular events as well.^25–27

A meta-analysis of primary prevention trials of aspirin was published in 2009.²² Although the relative risk reduction was similar in primary and secondary prevention, the absolute risk reduction in primary prevention was not nearly as great as in secondary prevention.

These findings are somewhat difficult to interpret, as the component trials included a wide spectrum of patients, ranging from healthy people with no symptoms and no known risk factors to those with limited risk factors. The trials were also performed over several decades during which primary prevention strategies were evolving. Additionally, most of the participants were middle-aged, nondiabetic men, so the results may not necessarily apply to people with diabetes, to women, or to the elderly. Thus, the pooled data in favor of aspirin for primary prevention may not be as broadly applicable to the general population as was once thought.

Aspirin for primary prevention in women

Guidelines for aspirin use in primary prevention were initially thought to be equally applicable to both sexes. However, concerns about the relatively low number of women participating in the studies and the possible mechanistic differences in aspirin efficacy in men vs women prompted further study.

A meta-analysis of randomized controlled trials found that aspirin was associated with a 12% relative reduction in the incidence of cardiovascular events in women and 14% in men. On the other hand, for stroke, the relative risk reduction was 17% in women, while men had no benefit.³²

Most of the women in this meta-analysis were participants in the Women’s Health Study, and they were at low baseline risk.²⁸ Although only about 10% of patients in this study were over age 65, this older group accounted for most of the benefit: these older women had a 26% risk reduction in major adverse cardiovascular events and 30% reduction in stroke.

Thus, for women, aspirin seems to become effective for primary prevention at an older age than in men, and the guidelines have been changed accordingly (Figure 1).

More women should be taking aspirin than actually are. For example, Rivera et al³³ found that only 41% of eligible women were receiving aspirin for primary prevention and 48% of eligible women were receiving it for secondary prevention.

People with diabetes

People with diabetes without overt cardiovascular disease are at higher risk of cardiovascular events than age- and sex-matched controls.³⁴ On the other hand, people with diabetes may be more prone to aspirin resistance and may not derive as much cardiovascular benefit from aspirin.

Early primary prevention studies included few people with diabetes. Subsequent meta-analyses of trials that used a wide range of aspirin doses found a relative risk reduction of 9%, which was not statistically significant.^9,35,36

But there is some evidence that people with diabetes, with³⁷ and without²² coronary disease, may be at higher inherent risk of bleeding than people without diabetes. Although aspirin may not necessarily increase the risk of bleeding in diabetic patients, recent data suggest no benefit in terms of a reduction in vascular events.³⁸

The balance of risk vs benefit for aspirin in this special population is not clear, although some argue that these patients should be treated somewhere on the spectrum of risk between primary and secondary prevention.

The US Preventive Services Task Force did not differentiate between people with or without diabetes in its 2009 guidelines for aspirin for primary prevention.⁸ However, the debate is reflected in a change in 2010 American College of Cardiology/American Diabetes Association guidelines regarding aspirin use in people with diabetes without known cardiovascular disease.³⁹ As opposed to earlier recommendations from these organizations in favor of aspirin for all people with diabetes regardless of 10-year risk, current recommendations advise low-dose aspirin (81–162 mg) for diabetic patients without known vascular disease who have a greater than 10% risk of a cardiovascular event and are not at increased risk of bleeding.

These changes were based on the findings of two trials: the Prevention and Progression of Arterial Disease and Diabetes Trial (POPADAD) and the Japanese Primary Prevention of Atherosclerosis With Aspirin for Diabetes (JPAD) study. These did not show a statistically significant benefit in prevention of cardiovascular events with aspirin.^29,30

After the new guidelines came out, a meta-analysis further bolstered its recommendations. ⁴⁰ In seven randomized clinical trials in 11,000 patients, the relative risk reduction was 9% with aspirin, which did not reach statistical significance.

The use of statins has been increasing, and this trend may have played a role in the marginal benefit of aspirin therapy in these recent studies. In the Japanese trial, approximately 25% of the patients were known to be using a statin; the percentage of statin use was not reported specifically in POPADAD, but both of these studies were published in 2008, when the proportion of diabetic patients taking a statin would be expected to be higher than in earlier primary prevention trials, which were performed primarily in the 1990s. Thus, the beneficial effects of statins may have somewhat diluted the risk reduction attributable to aspirin.

Trials under way in patients with diabetes

The evolving and somewhat conflicting guidelines highlight the need for further study in patients with diabetes. To address this area, two trials are in progress: the Aspirin and Simvastatin Combination for Cardiovascular Events Prevention Trial in Diabetes (ACCEPT-D) and A Study of Cardiovascular Events in Diabetes (ASCEND).^41,42

ACCEPT-D is testing low-dose aspirin (100 mg daily) in diabetic patients who are also on simvastatin. This study also includes prespecified subgroups stratified by sex, age, and baseline lipid levels.

The ASCEND trial will use the same aspirin dose as ACCEPT-D, with a target enrollment of 10,000 patients with diabetes without known vascular disease.

More frequent dosing for people with diabetes?

Although not supported by current guidelines, recent work has suggested that people with diabetes may need more-frequent dosing of aspirin.⁴³ This topic warrants further investigation.

Aspirin as primary prevention in elderly patients

The incidence of cardiovascular events increases with age³⁷—but so does the incidence of gastrointestinal bleeding.⁴⁴ Upper gastrointestinal bleeding is especially worrisome in the elderly, in whom the estimated case-fatality rate is high.¹² Assessment of risk and benefit is particularly important in patients over age 65 without known coronary disease.

Uncertainty about aspirin use in this population is reflected in the most recent US Preventive Services Task Force guidelines, which do not advocate either for or against regular aspirin use for primary prevention in those over the age of 80.

Data on this topic from clinical trials are limited. The Antithrombotic Trialists’ Collaboration (2009) found that although age is associated with a risk of major coronary events similar to that of other traditional risk factors such as diabetes, hypertension, and tobacco use, older age is also associated with the highest risk of major extracranial bleeding.²²

Because of the lack of data in this population, several studies are currently under way. The Aspirin in Reducing Events in the Elderly (ASPREE) trial is studying 100 mg daily in nondiabetic patients without known cardiovascular disease who are age 70 and older.⁴⁵ An additional trial will study patients age 60 to 85 with concurrent diagnoses of hypertension, hyperlipidemia, or diabetes and will test the same aspirin dose as in ASPREE.⁴⁶ These trials should provide further insight into the safety and efficacy of aspirin for primary prevention in the elderly.

FUTURE DIRECTIONS

Aspirin remains a cornerstone of therapy in patients with cardiovascular disease and in secondary prevention of adverse cardiovascular events, but its role in primary prevention remains under scrutiny. Recommendations have evolved to reflect emerging data in special populations, and an algorithm based on Framingham risk assessment in men for myocardial infarction and ischemic stroke assessment in women for assessing appropriateness of aspirin therapy based on currently available guidelines is presented in Figure 1.^6,8,47–49 Targeted studies have advanced our understanding of aspirin use in women, and future studies in people with diabetes and in the elderly should provide further insight into the role of aspirin for primary prevention in these specific groups as well.

Additionally, the range of doses used in clinical studies has propagated the general misperception that higher doses of aspirin are more efficacious. Future studies should continue to use lower doses of aspirin to minimize bleeding risk with an added focus on re-examining its net benefit in the modern era of increasing statin use, which may reduce the absolute risk reduction attributable to aspirin.

One particular area of debate is whether enteric coating can result in functional aspirin resistance. Grosser et al⁵⁰ found that sustained aspirin resistance was rare, and “pseudoresistance” was related to the use of a single enteric-coated aspirin instead of immediate-release aspirin in people who had not been taking aspirin up to then. This complements an earlier study, which found that enteric-coated aspirin had an appropriate effect when given for 7 days.⁵¹ Therefore, for patients who have not been taking aspirin, the first dose should always be immediate-release, not enteric coated.

SHOULD OUR PATIENT RECEIVE ASPIRIN?

The patient we described at the beginning of this article has several risk factors—hypertension, dyslipidemia, left ventricular hypertrophy, and smoking—but no known cardiovascular disease as yet. Her risk of an adverse cardiovascular event appears moderate. However, her 10-year risk of stroke by the Framingham risk calculation is 10%, which would qualify her for aspirin for primary prevention. Of particular note is that the significance of left ventricular hypertrophy as a risk factor for stroke in women is higher than in men and in our case accounts for half of this patient’s risk.

We should explain to the patient that the anticipated benefits of aspirin for stroke prevention outweigh bleeding risks, and thus aspirin therapy would be recommended. However, with her elevated LDL-cholesterol, she may benefit from a statin, which could lessen the relative risk reduction from additional aspirin use.

A 57-year-old woman with no history of cardiovascular disease comes to the clinic for her annual evaluation. She does not have diabetes mellitus, but she does have hypertension and chronic osteoarthritis, currently treated with acetaminophen. Additionally, she admits to active tobacco use. Her systolic blood pressure is 130 mm Hg on therapy with hydrochlorothiazide. Her electrocardiogram demonstrates left ventricular hypertrophy. Her low-density lipoprotein (LDL) cholesterol level is 140 mg/dL, and her high-density lipoprotein (HDL) cholesterol level is 50 mg/dL. Should this patient be started on aspirin therapy?

Acetylsalicylic acid (aspirin) is an analgesic, antipyretic, and anti-inflammatory agent, but its more prominent use today is as an antithrombotic agent to treat or prevent cardiovascular events. Its antithrombotic properties are due to its effects on the enzyme cyclooxygenase. However, cyclooxygenase is also involved in regulation of the gastric mucosa, and so aspirin increases the risk of gastrointestinal bleeding.

Approximately 50 million people take aspirin on a daily basis to treat or prevent cardiovascular disease.¹ Of these, at least half are taking more than 100 mg per day,² reflecting the general belief that, for aspirin dosage, “more is better”—which is not true.

Additionally, recommendations about the use of aspirin were based on studies that included relatively few members of several important subgroups, such as people with diabetes without known cardiovascular disease, women, and the elderly, and thus may not reflect appropriate indications and dosages for these groups.

Here, we examine the literature, outline an individualized approach to aspirin therapy, and highlight areas for future study.

HISTORY OF ASPIRIN USE IN CARDIOVASCULAR DISEASE

• 1700s—Willow bark is used as an analgesic.
• 1897—Synthetic aspirin is developed as an antipyretic and anti-inflammatory agent.
• 1974—First landmark trial of aspirin for secondary prevention of myocardial infarction.³
• 1982—Nobel Prize awarded for discovery of aspirin mechanism.
• 1985—US Food and Drug Administration approves aspirin for the treatment and secondary prevention of acute myocardial infarction.
• 1998—The Second International Study of Infarct Survival (ISIS-2) finds that giving aspirin to patients with myocardial infarction within 24 hours of presentation leads to a significant reduction in vascular deaths.⁴

Ongoing uncertainties

Aspirin now carries a class I indication for all patients with suspected myocardial infarction. Since there are an estimated 600,000 new coronary events and 325,000 recurrent ischemic events per year in the United States,⁵ the need for aspirin will continue to remain great. It is also approved to prevent and treat stroke and in patients with unstable angina.

However, questions continue to emerge about aspirin’s dosing and appropriate use in specific populations. The initial prevention trials used a wide range of doses and, as mentioned, included few women, few people with diabetes, and few elderly people. The uncertainties are especially pertinent for patients without known vascular disease, in whom the absolute risk reduction is much less, making the assessment of bleeding risk particularly important. Furthermore, the absolute risk-to-benefit assessment may be different in certain populations.

Guidelines on the use of aspirin to prevent cardiovascular disease (Table 1)^6–10 have evolved to take into account these possible disparities, and studies are taking place to further investigate aspirin use in these groups.

ASPIRIN AND GASTROINTESTINAL BLEEDING

Aspirin’s association with bleeding, particularly gastrointestinal bleeding, was recognized early as a use-limiting side effect. With or without aspirin, gastrointestinal bleeding is a common cause of morbidity and death, with an incidence of approximately 100 per 100,000 bleeding episodes in adults per year for upper gastrointestinal bleeding and 20 to 30 per 100,000 per year for lower gastrointestinal bleeding.^11,12

The standard dosage (ie, 325 mg/day) is associated with a significantly higher risk of gastrointestinal bleeding (including fatal bleeds) than is 75 mg.¹³ However, even with lower doses, the risk of gastrointestinal bleeding is estimated to be twice as high as with no aspirin.¹⁴

And here is the irony: studies have shown that higher doses of aspirin offer no advantage in preventing thrombotic events compared with lower doses.¹⁵ For example, the Clopidogrel Optimal Loading Dose Usage to Reduce Recurrent Events/Organization to Assess Strategies for Ischemic Stroke Syndromes study reported a higher rate of gastrointestinal bleeding with standard-dose aspirin therapy than with low-dose aspirin, with no additional cardiovascular benefit with the higher dose.¹⁶

Furthermore, several other risk factors increase the risk of gastrointestinal bleeding with aspirin use (Table 2). These risk factors are common in the general population but were not necessarily represented in participants in clinical trials. Thus, estimates of risk based on trial data most likely underestimate actual risk in the general population, and therefore, the individual patient’s risk of gastrointestinal bleeding, based on these and other factors, needs to be taken into consideration.

ASPIRIN IN PATIENTS WITH CORONARY ARTERY DISEASE

Randomized clinical trials have validated the benefits of aspirin in secondary prevention of cardiovascular events in patients who have had a myocardial infarction. Patients with coronary disease who withdraw from aspirin therapy or otherwise do not adhere to it have a risk of cardiovascular events three times higher than those who stay with it.¹⁷

Despite the strong data, however, several issues and questions remain about the use of aspirin for secondary prevention.

Bleeding risk must be considered, since gastrointestinal bleeding is associated with a higher risk of death and myocardial infarction in patients with cardiovascular disease.¹⁸ Many patients with coronary disease are on more than one antiplatelet or anticoagulant therapy for concomitant conditions such as atrial fibrillation or because they underwent a percutaneous intervention, which further increases the risk of bleeding.

This bleeding risk is reflected in changes in the most recent recommendations for aspirin dosing after percutaneous coronary intervention. Earlier guidelines advocated use of either 162 or 325 mg after the procedure. However, the most recent update (in 2011) now supports 81 mg for maintenance dosing after intervention.⁷

Patients with coronary disease but without prior myocardial infarction or intervention. Current guidelines recommend 75 to 162 mg of aspirin in all patients with coronary artery disease.⁶ However, this group is diverse and includes patients who have undergone percutaneous coronary intervention, patients with chronic stable angina, and patients with asymptomatic coronary artery disease found on imaging studies. The magnitude of benefit is not clear for those who have no symptoms or who have stable angina.

Most of the evidence supporting aspirin use in chronic angina came from a single trial in Sweden, in which 2,000 patients with chronic stable angina were given either 75 mg daily or placebo. Those who received aspirin had a 34% lower rate of myocardial infarction and sudden death.¹⁹

A substudy of the Physicians’ Health Study, with fewer patients, also noted a significant reduction in the rate of first myocardial infarction. The dose of aspirin in this study was 325 mg every other day.²⁰

In the Women’s Health Initiative Observational Study, 70% of women with stable cardiovascular disease taking aspirin were taking 325 mg daily.²¹ This study demonstrated a significant reduction in the cardiovascular mortality rate, which supports current guidelines, and found no difference in outcomes with doses of 81 mg compared with 325 mg.²¹ This again corroborates that low-dose aspirin is preferential to standard-dose aspirin in women with cardiovascular disease.

These findings have not been validated in larger prospective trials. Thus, current guidelines for aspirin use may reflect extrapolation of aspirin benefit from higher-risk patients to lower-risk patients.

Nevertheless, although the debate continues, it has generally been accepted that in patients who are at high risk of vascular disease or who have had a myocardial infarction, the benefits of aspirin—a 20% relative reduction in vascular events²²—clearly outweigh the risks.

ASPIRIN FOR PRIMARY PREVENTION

Assessing risk vs benefit is more complex when considering populations without known cardiovascular disease.

Only a few studies have specifically evaluated the use of aspirin for primary prevention (Table 3).^23–31 The initial trials were in male physicians in the United Kingdom and the United States in the late 1980s and had somewhat conflicting results. A British study did not find a significant reduction in myocardial infarction,²³ but the US Physician’s Health Study study did: the relative risk was 0.56 (95% confidence interval 0.45–0.70, P < .00001).²⁴ The US study had more than four times the number of participants, used different dosing (325 mg every other day compared with 500 or 300 mg daily in the British study), and had a higher rate of compliance.

Several studies over the next decade demonstrated variable but significant reductions in cardiovascular events as well.^25–27

A meta-analysis of primary prevention trials of aspirin was published in 2009.²² Although the relative risk reduction was similar in primary and secondary prevention, the absolute risk reduction in primary prevention was not nearly as great as in secondary prevention.

These findings are somewhat difficult to interpret, as the component trials included a wide spectrum of patients, ranging from healthy people with no symptoms and no known risk factors to those with limited risk factors. The trials were also performed over several decades during which primary prevention strategies were evolving. Additionally, most of the participants were middle-aged, nondiabetic men, so the results may not necessarily apply to people with diabetes, to women, or to the elderly. Thus, the pooled data in favor of aspirin for primary prevention may not be as broadly applicable to the general population as was once thought.

Aspirin for primary prevention in women

Guidelines for aspirin use in primary prevention were initially thought to be equally applicable to both sexes. However, concerns about the relatively low number of women participating in the studies and the possible mechanistic differences in aspirin efficacy in men vs women prompted further study.

A meta-analysis of randomized controlled trials found that aspirin was associated with a 12% relative reduction in the incidence of cardiovascular events in women and 14% in men. On the other hand, for stroke, the relative risk reduction was 17% in women, while men had no benefit.³²

Most of the women in this meta-analysis were participants in the Women’s Health Study, and they were at low baseline risk.²⁸ Although only about 10% of patients in this study were over age 65, this older group accounted for most of the benefit: these older women had a 26% risk reduction in major adverse cardiovascular events and 30% reduction in stroke.

Thus, for women, aspirin seems to become effective for primary prevention at an older age than in men, and the guidelines have been changed accordingly (Figure 1).

More women should be taking aspirin than actually are. For example, Rivera et al³³ found that only 41% of eligible women were receiving aspirin for primary prevention and 48% of eligible women were receiving it for secondary prevention.

People with diabetes

People with diabetes without overt cardiovascular disease are at higher risk of cardiovascular events than age- and sex-matched controls.³⁴ On the other hand, people with diabetes may be more prone to aspirin resistance and may not derive as much cardiovascular benefit from aspirin.

Early primary prevention studies included few people with diabetes. Subsequent meta-analyses of trials that used a wide range of aspirin doses found a relative risk reduction of 9%, which was not statistically significant.^9,35,36

But there is some evidence that people with diabetes, with³⁷ and without²² coronary disease, may be at higher inherent risk of bleeding than people without diabetes. Although aspirin may not necessarily increase the risk of bleeding in diabetic patients, recent data suggest no benefit in terms of a reduction in vascular events.³⁸

The balance of risk vs benefit for aspirin in this special population is not clear, although some argue that these patients should be treated somewhere on the spectrum of risk between primary and secondary prevention.

The US Preventive Services Task Force did not differentiate between people with or without diabetes in its 2009 guidelines for aspirin for primary prevention.⁸ However, the debate is reflected in a change in 2010 American College of Cardiology/American Diabetes Association guidelines regarding aspirin use in people with diabetes without known cardiovascular disease.³⁹ As opposed to earlier recommendations from these organizations in favor of aspirin for all people with diabetes regardless of 10-year risk, current recommendations advise low-dose aspirin (81–162 mg) for diabetic patients without known vascular disease who have a greater than 10% risk of a cardiovascular event and are not at increased risk of bleeding.

These changes were based on the findings of two trials: the Prevention and Progression of Arterial Disease and Diabetes Trial (POPADAD) and the Japanese Primary Prevention of Atherosclerosis With Aspirin for Diabetes (JPAD) study. These did not show a statistically significant benefit in prevention of cardiovascular events with aspirin.^29,30

After the new guidelines came out, a meta-analysis further bolstered its recommendations. ⁴⁰ In seven randomized clinical trials in 11,000 patients, the relative risk reduction was 9% with aspirin, which did not reach statistical significance.

The use of statins has been increasing, and this trend may have played a role in the marginal benefit of aspirin therapy in these recent studies. In the Japanese trial, approximately 25% of the patients were known to be using a statin; the percentage of statin use was not reported specifically in POPADAD, but both of these studies were published in 2008, when the proportion of diabetic patients taking a statin would be expected to be higher than in earlier primary prevention trials, which were performed primarily in the 1990s. Thus, the beneficial effects of statins may have somewhat diluted the risk reduction attributable to aspirin.

Trials under way in patients with diabetes

The evolving and somewhat conflicting guidelines highlight the need for further study in patients with diabetes. To address this area, two trials are in progress: the Aspirin and Simvastatin Combination for Cardiovascular Events Prevention Trial in Diabetes (ACCEPT-D) and A Study of Cardiovascular Events in Diabetes (ASCEND).^41,42

ACCEPT-D is testing low-dose aspirin (100 mg daily) in diabetic patients who are also on simvastatin. This study also includes prespecified subgroups stratified by sex, age, and baseline lipid levels.

The ASCEND trial will use the same aspirin dose as ACCEPT-D, with a target enrollment of 10,000 patients with diabetes without known vascular disease.

More frequent dosing for people with diabetes?

Although not supported by current guidelines, recent work has suggested that people with diabetes may need more-frequent dosing of aspirin.⁴³ This topic warrants further investigation.

Aspirin as primary prevention in elderly patients

The incidence of cardiovascular events increases with age³⁷—but so does the incidence of gastrointestinal bleeding.⁴⁴ Upper gastrointestinal bleeding is especially worrisome in the elderly, in whom the estimated case-fatality rate is high.¹² Assessment of risk and benefit is particularly important in patients over age 65 without known coronary disease.

Uncertainty about aspirin use in this population is reflected in the most recent US Preventive Services Task Force guidelines, which do not advocate either for or against regular aspirin use for primary prevention in those over the age of 80.

Data on this topic from clinical trials are limited. The Antithrombotic Trialists’ Collaboration (2009) found that although age is associated with a risk of major coronary events similar to that of other traditional risk factors such as diabetes, hypertension, and tobacco use, older age is also associated with the highest risk of major extracranial bleeding.²²

Because of the lack of data in this population, several studies are currently under way. The Aspirin in Reducing Events in the Elderly (ASPREE) trial is studying 100 mg daily in nondiabetic patients without known cardiovascular disease who are age 70 and older.⁴⁵ An additional trial will study patients age 60 to 85 with concurrent diagnoses of hypertension, hyperlipidemia, or diabetes and will test the same aspirin dose as in ASPREE.⁴⁶ These trials should provide further insight into the safety and efficacy of aspirin for primary prevention in the elderly.

FUTURE DIRECTIONS

Aspirin remains a cornerstone of therapy in patients with cardiovascular disease and in secondary prevention of adverse cardiovascular events, but its role in primary prevention remains under scrutiny. Recommendations have evolved to reflect emerging data in special populations, and an algorithm based on Framingham risk assessment in men for myocardial infarction and ischemic stroke assessment in women for assessing appropriateness of aspirin therapy based on currently available guidelines is presented in Figure 1.^6,8,47–49 Targeted studies have advanced our understanding of aspirin use in women, and future studies in people with diabetes and in the elderly should provide further insight into the role of aspirin for primary prevention in these specific groups as well.

Additionally, the range of doses used in clinical studies has propagated the general misperception that higher doses of aspirin are more efficacious. Future studies should continue to use lower doses of aspirin to minimize bleeding risk with an added focus on re-examining its net benefit in the modern era of increasing statin use, which may reduce the absolute risk reduction attributable to aspirin.

One particular area of debate is whether enteric coating can result in functional aspirin resistance. Grosser et al⁵⁰ found that sustained aspirin resistance was rare, and “pseudoresistance” was related to the use of a single enteric-coated aspirin instead of immediate-release aspirin in people who had not been taking aspirin up to then. This complements an earlier study, which found that enteric-coated aspirin had an appropriate effect when given for 7 days.⁵¹ Therefore, for patients who have not been taking aspirin, the first dose should always be immediate-release, not enteric coated.

SHOULD OUR PATIENT RECEIVE ASPIRIN?

The patient we described at the beginning of this article has several risk factors—hypertension, dyslipidemia, left ventricular hypertrophy, and smoking—but no known cardiovascular disease as yet. Her risk of an adverse cardiovascular event appears moderate. However, her 10-year risk of stroke by the Framingham risk calculation is 10%, which would qualify her for aspirin for primary prevention. Of particular note is that the significance of left ventricular hypertrophy as a risk factor for stroke in women is higher than in men and in our case accounts for half of this patient’s risk.

We should explain to the patient that the anticipated benefits of aspirin for stroke prevention outweigh bleeding risks, and thus aspirin therapy would be recommended. However, with her elevated LDL-cholesterol, she may benefit from a statin, which could lessen the relative risk reduction from additional aspirin use.