Healthy Lifestyle Factors in the Primary Prevention of Coronary Heart Disease Among Men

Major risk factors for coronary heart disease (CHD), such as hypercholesterolemia, hypertension, and smoking, are modifiable.¹ Pharmacological therapies, including lipid-lowering and antihypertensive medications, are efficacious in lowering some of these risk factors,^2,3 typically reducing cardiovascular disease risk by 20% to 30%.^2,4 Although pharmacological agents successfully reduce coronary events, the overall reduction in risk is relatively modest and could be greatly improved by the addition of lifestyle modifications.

Healthy diet and other lifestyle practices, including not smoking, maintaining a healthy weight, and exercising daily, also improve these clinical risk factors as well as reduce inflammation, homocysteine, glucose intolerance, and arrhythmias.^5–11 Individually, these modifiable lifestyle factors are associated with lower risk of CHD,¹² but a combination of healthy lifestyle choices has a greater impact than any single lifestyle factor. A healthy lifestyle has been associated with lower risk of incident coronary events among middle-aged women¹³ and lower risk of CHD mortality in a population of elderly men and women.¹⁴ Little is known about the relation between combined healthy lifestyle choices and incident CHD among middle-aged and older men, especially those who are reducing their risk by using drugs for hypertension or hypercholesterolemia.

We estimated the burden of CHD that could potentially be avoided through a healthy lifestyle among highly educated middle-aged and older US men. We hypothesized that among men taking medication for hypertension or hypercholesterolemia, these healthy lifestyle characteristics would still remain associated with lower CHD risk. Finally, we examined the association between changes in lifestyle during middle and older ages and risk of CHD among these men.

Methods

Study Population

The Health Professionals Follow-up Study (HPFS) is a cohort of 51 529 US male health professionals, ages 40 to 75 years (mean age, 53) at baseline in 1986.¹⁵ Participants have provided information on current medical conditions and lifestyle factors biennially starting in 1986 through the use of self-administered questionnaires. We excluded men with missing or implausible nutrient intakes at baseline (≥70 items left blank or estimated total energy intake ≤800 or ≥4200 kcal) and those with a history of chronic disease. The final population for this analysis was 42 847. The Institutional Review Board of the Harvard School of Public Health approved the study protocol.

Exposure and Covariate Ascertainment

Self-reported height and family history of myocardial infarction (MI) before 60 years of age were ascertained on the baseline questionnaire. Data on weight, medication use, current smoking status, and physician diagnosis of hypertension and hypercholesterolemia were assessed every 2 years. We calculated body mass index (BMI [kg/m²]) by using self-reported measures of height and weight, which have been validated in a subset of this cohort.¹⁶

Participants reported the average time engaged in specific activities biennially, using an activity questionnaire previously validated within a subset of this cohort.¹⁷ We calculated hours per week of moderate-to-vigorous activity requiring at least 4 metabolic equivalents. We included walking at a brisk pace (≥3 mph), jogging, running, bicycling, swimming, tennis, squash, racquetball, rowing, and calisthenics.

We assessed dietary information by using a 131-item, semiquantitative food frequency questionnaire (FFQ), administered every 4 years. The reproducibility and validity of these FFQs are high when compared with multiple 1-week diet records and biochemical markers.^18–20

We calculated a summary dietary score based on the Alternate Healthy Eating Index (AHEI).²¹ The AHEI is a modification of the Healthy Eating Index (HEI), created by the US Department of Agriculture to assess how well the US population met dietary recommendations based on the Food Guide Pyramid and the Dietary Guidelines for Americans.²² The AHEI, which targets foods and nutrients associated with lower risk of chronic disease, better predicts chronic disease risk than do other measures of diet quality, including the HEI.^21,23 Although alcohol was part of the original AHEI, for the present study we considered it a separate lifestyle factor. With the exception of multivitamin use, the AHEI components were given a score ranging from 0 to 10, where 10 signified optimal dietary behavior. The 7 specific components (criteria for minimum, maximum scores) were percent energy from trans fat (≥4%, ≤0.5%), ratio of polyunsaturated:saturated fat (≤0.1, ≥1), and chicken plus fish:red meat (0, ≥4; a small percentage of vegetarians were given a score of 10), daily servings of fruit (0, ≥4), vegetables (0, ≥5), vegetable proteins (legumes, tofu, and soy products) (0, ≥1), and grams of cereal fiber (0, ≥15). The eighth component, multivitamin use for ≥5 years, was dichotomous, to avoid overweighting this component (yes=7.5, no=2.5 points). Our diet score ranged from 2.5 (worst) to 77.5 (best).

For each lifestyle factor (smoking, BMI, exercise, diet score, and alcohol), we created a binary low-risk variable, where the men received a 1 if they met the criteria for low risk and a 0 if otherwise. The a priori definition of low risk was based on the current literature and recommended guidelines but also on levels realistically obtainable within the general population. For smoking, we defined low risk as not currently smoking. Because we wanted to consider only modifiable factors, we included former smokers in our low-risk category, as current smokers cannot attain the status of “never smoker.” The risk of CHD among former smokers declines after smoking cessation, approximating the risk of those who have never smoked after 10 to 14 years.²⁴ For exercise, at least 30 min/d of moderate-to-vigorous intensity was considered low risk, on the basis of current guidelines.²⁵ Optimal body weight was defined as a BMI of <25 kg/m², the standard World Health Organization cutoff for healthy weight. We considered average daily alcohol intake of 5 to 30 g as low risk. This is consistent with guidelines for moderate alcohol consumption,²⁶ although higher amounts of alcohol intake are associated with lower risk of CHD.²⁷ For diet, low risk was defined as a diet score in the top 40% of the cohort distribution, as there are no recommendations established for the AHEI. The average diet score among the men with low-risk diets was 50.3. An average low-risk diet of a long-term multivitamin user would consist of approximately 3 servings of vegetables, 2.5 servings of fruit, 0.5 servings of nuts, 9 g of cereal fiber, 1.8% of calories from trans fat, 2.5 servings of chicken and fish for 1 serving of red meat, and a polyunsaturated:saturated fat ratio of 0.6.

We calculated a healthy lifestyle score by summing the total number of lifestyle factors for which the men were at low risk. The men could obtain a healthy lifestyle score from 0 (least healthy) to 5 (most healthy).

Outcome Ascertainment

The outcome for this analysis was incident CHD, defined as nonfatal MI or fatal CHD, occurring between the return of the baseline questionnaire in 1986 and January 31, 2002. Confirmed MIs were defined according to World Health Organization criteria,²⁸ and, when available, cardiac-specific troponin levels were used. If an individual was admitted to the hospital for an MI but medical records were unavailable, the infarction was considered probable. Confirmed deaths were those caused by MI, according to autopsy or hospital records, or if CHD was listed as the cause of death and previous evidence of CHD was obtained. Deaths were considered probable if the cause of death was CHD without confirmation of previous CHD or those listed as sudden deaths with no other potential cause. We used both confirmed and probable cases, as results were similar when we excluded probable cases (17% of cases).

Statistical Analysis

We used the cumulative average of the AHEI-based diet score from repeated dietary assessments to represent long-term dietary information and minimize within-person variation.²⁹ For example, the diet score from the 1986 FFQ was used to estimate CHD risk between 1986 and 1990, whereas an average of the 1986 and 1990 diet score was used to predict CHD risk occurring between 1990 and 1994. The average of the 1986, 1990, and 1994 diet scores was used to estimate disease risk between 1994 and 1998, and so on. Because diagnosis of diabetes, angina, hypertension, hypercholesterolemia, or revascularization surgery may lead to changes in diet, we stopped updating dietary information after new diagnoses of these outcomes during follow-up. Nondietary factors were updated every 2 years.

For individuals missing information on any lifestyle factor, we carried forward information from the previous questionnaire, when available. Men with residual missing values were placed in the high-risk category, to give the most conservative estimate. Less than 1% of values were missing for each lifestyle factor; among the current smokers, <2% of men were missing information on number of cigarettes.

Each individual contributed person-time from the return of the 1986 questionnaire until the date of first coronary event, date of diagnosis of cancer or stroke, death, or January 31, 2002, whichever came first. Relative risks (RRs) and 95% confidence intervals (CIs) were estimated by using Cox proportional hazards models adjusted for age, calendar year, parental history of MI, history of hypertension, history of hypercholesterolemia, aspirin use, and antihypertensive medication use. Further adjustment for lipid-lowering medications did not appreciably alter the results.

We calculated the population-attributable risk³⁰ and 95% CI³¹ to estimate the proportion of cases within the population that could have been avoided had all the men adhered to the low-risk lifestyle practices, assuming a causal relation between the low-risk lifestyle practices and risk of CHD. To calculate the population-attributable risk, we used Cox proportional hazards models to calculate relative risks of CHD, comparing men at low risk in a specific combination of lifestyle factors with the rest of the men in the population.³²

To assess whether modifiable lifestyle factors were associated with lower risk among men already undergoing drug therapy, we examined the relation between healthy lifestyle score and risk of CHD separately among subgroups of users and nonusers of lipid-lowering or antihypertensive medications. We defined medication use as current use, which was updated every 2 years.

We estimated changes in lifestyle and risk of CHD by modeling the difference between the attained healthy lifestyle score at follow-up and the healthy lifestyle score at baseline, controlling for baseline score. For this analysis, we included only men who completed the baseline questionnaire and at least one questionnaire during the follow-up period (n=33 759).

The authors had full access to the data and take responsibility for its integrity. All authors have read and agreed to the manuscript as written.

Results

After 16 years of follow-up, we documented 2183 incident coronary events. The frequencies of each lifestyle factor within this population are shown in Table 1. For each lifestyle factor, we observed an inverse association with risk of CHD, which remained significant when all 5 lifestyle factors were included in the same model (Table 1). When categorized as a dichotomous variable, being at low risk for each lifestyle factor was still significantly and independently associated with lower risk.

Overall, the healthy lifestyle score was significantly inversely associated with risk of CHD (P for trend, <0.0001) (Figure 1). Men at low risk for ≥1 lifestyle factor had a significantly lower risk of CHD, compared with men at low risk for none of the 5 factors. Men at low risk for all 5 factors had the lowest risk.

• Download figure
• Download PowerPoint

Men in this population reported use of lipid-lowering or antihypertensive medication use for 21% of the person-time during follow-up. Among these medication users (and also among nonusers), the healthy lifestyle score was significantly inversely associated with risk of CHD (Figure 2). Results were not appreciably different when we excluded men who were diagnosed with hypertension or hypercholesterolemia during follow-up but reported no medication use (data not shown).

• Download figure
• Download PowerPoint

We calculated the population-attributable risk for men who adhered to a combination of low-risk lifestyle practices (Table 2). Men in the low-risk group for all 5 lifestyle practices, only 4% of the population, had an RR of 0.37 (95% CI: 0.26, 0.53), compared with the remaining men in the population. This translates to a population-attributable risk of 62% (95% CI: 49%, 74%), suggesting that a majority of coronary disease in this population may be attributed to poor adherence to a healthy lifestyle.

Among medication users, the population-attributable risk for adhering to all 5 low-risk lifestyle practices was 57% (95% CI: 32%, 79%) (Table 3). We found similar associations among antihypertensive and lipid-lowering medications users separately albeit with wider confidence intervals. The population-attributable risk for adhering to all 5 low-risk lifestyle practices was much higher among middle-aged men (<65 years old) (population-attributable risk: 79%; 95% CI: 61%, 90%) than among men 65 years or older (population-attributable risk: 47%; 95% CI: 27%, 68%) (Table 3).

Because of the long follow-up period and the repeated assessment of lifestyle characteristics over time, we were able to examine the CHD risk associated with changes in the healthy lifestyle score. Among the men included in the change in lifestyle analysis, there were 1583 incident coronary events. Men who adopted at least 2 new healthy characteristics during follow-up had a relative risk of 0.73 (95% CI: 0.57, 0.93) compared with men who did not change, controlling for the number at baseline (Table 4). On the other hand, men who reduced their healthy lifestyle score by 2 or more factors had a relative risk of 1.48 (95% CI: 1.15, 1.88), compared with men who did not change.

Discussion

In this population of male health professionals, 62% of all coronary events may have been avoided had all men adhered to a low-risk lifestyle of not smoking, exercising regularly, eating prudently, consuming alcohol in moderation, and maintaining a healthy weight. Equally important, we found that this combination of healthy lifestyle characteristics was strongly inversely associated with risk even among men taking medication for coronary risk factors. Further, we found that middle-aged and older men who adopted additional healthy lifestyle practices over time further lowered their risk of coronary disease.

The therapeutic benefits of antihypertensive and lipid-lowering drugs are well documented,^2,4 but coronary disease remains elevated in this population. The addition of healthy behaviors, such as a prudent diet and regular exercise, to either antihypertensive or lipid-lowering drug therapy in other populations reduced risk factors, such as systolic and diastolic blood pressure^33,34 and low-density lipoprotein cholesterol,³⁵ as well as the inflammatory biomarker C-reactive protein,³⁶ compared with drug therapy alone.

In our cohort, the population-attributable risk for men at low risk for all 5 lifestyle factors was 57% among the users of cardiovascular medications. This implies that more than half of the cases of CHD among the medication users in this population may have been avoided by adherence to low-risk lifestyle practices, in addition to the benefits seen from medication use. Cardiovascular medications should be used as an adjunct to, not just a replacement for, healthy lifestyle practices, especially in the setting of primary prevention.

This study shows the extent to which healthy lifestyle changes may lower the risk of CHD, even during middle age or later in life. We probably underestimated the true benefit of healthy changes, as the men who were the healthiest (those with 4 or 5 low-risk lifestyle factors) could not increase their score by 2 factors. Furthermore, we could not capture the benefit among men who met the threshold for low risk but continued to improve their lifestyle. This was an observational study, not a clinical trial; thus, these men made changes of their own free will. Only 4% of the men adopted all healthy practices on their own; however, whether a greater number of these men would follow a low-risk regimen if given as a direct intervention cannot be determined from our data.

The population-attributable risk resulting from low-risk lifestyle practices within this population of men was lower than expected. However, when we excluded men >65 years of age, the results were similar to those seen among middle-aged women from the Nurses’ Health Study¹³ as well as from studies of CHD mortality conducted among younger and middle-aged men and women, using clinical definitions of low risk.^37,38 Even among older men (≥65 years), healthy lifestyle factors were associated with significantly lower risk. In an older Mediterranean population (70 to 90 years), risk of CHD mortality was 65% lower among participants who committed to 4 healthy lifestyle choices.¹⁴

Overall, these health professionals were at lower risk of developing CHD than men in the general population. The rate of incident CHD in this population (2.86 events/1000 person-years) was almost half of the age-standardized rate seen in the Atherosclerosis Risk in Communities (ARIC) study population (5.53 events/1000 person-years³⁹), the participants of which were selected to be representative of adults in 4 communities across the United States.⁴⁰ Summary results from the National Health Interview Survey^41–44 indicate that fewer men in the United States ≥45 years of age met our low-risk criteria for exercise (16% versus 29% in our cohort), optimal BMI (37% versus 43%), absence of current smoking (84% versus 91%), and moderate alcohol consumption (19% versus 40%). Nationally representative data needed to calculate the AHEI-based diet score are not available. The average HEI score of 68.3, within this cohort,⁴⁵ is slightly higher than the mean HEI score of 65.2 among US men 51 years or older.⁴⁶ Even though the men in our cohort were healthier than men in the United States, only 4% of men met the low-risk criteria for all 5 characteristics. We still conclude that a majority of cases of CHD among this low-risk population of men may have been prevented through a healthy lifestyle.

Although the accuracy of self-reported data within this population of motivated and educated health professionals has been well documented,^16–19 measurement error in self-reported variables is inevitable. Because of the prospective design of the study, this misclassification should be nondifferential with respect to disease status and would attenuate the true relative risk. A randomized clinical trial is ideal to establish causality of these lifestyle factors on risk of CHD. For some of these lifestyle factors, such as smoking and alcohol consumption, a clinical trial may not be ethical. Short-term trials on diet, physical activity, and weight loss have shown substantial reduction in coronary risk factors as well as coronary events.^47,48 In our analysis, we estimated the impact of multiple lifestyle factors simultaneously, which would be difficult if not impossible to study in a trial with 16 years of follow-up. The observational design and repeated measures of diet and lifestyle allowed us to assess the consequences of changes in multiple lifestyle factors, which generally occur in free-living populations.

In conclusion, a healthy lifestyle plays an important role in the primary prevention of CHD in middle-aged and older men, even among men on antihypertensive or lipid-lowering medication. A healthy lifestyle can be an effective, nonpharmacological approach to reducing coronary heart disease among men.

Footnotes