The Joint Effects of Cardiorespiratory Fitness and Adiposity on Mortality Risk in Men With Hypertension

Abstract

Background

Whether higher cardiorespiratory fitness (CRF) attenuates the mortality risk associated with higher adiposity in adults with hypertension (HTN) is poorly understood.

Methods

Participants were 13,155 men (mean age, 47.7 (s.d., 9.9) years) who completed a baseline health examination and maximal treadmill exercise test during 1974–2003. All men had HTN at baseline based on resting systolic blood pressure of ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg. CRF was quantified as the duration of a symptom-limited maximal treadmill exercise test, and was grouped for analysis as low (lowest 20%), moderate (middle 40%), and high (upper 40%). Distributions of body mass index (BMI), waist circumference (WC), and percent body fat (%BF) were grouped according to standard clinical guidelines.

Results

During a mean follow-up of 12 years, 883 deaths (355 cardiovascular disease (CVD)) were recorded. Multivariate hazard ratios (HRs) (95% confidence interval) for all-cause mortality, using low-fitness as the reference group, were 0.58 (0.48–0.69) and 0.43 (0.35–0.54) for moderate-fit and high-fit groups, respectively. We observed a similar pattern for CVD mortality. High-fit/obese men had no greater risk of all-cause (1.59 (0.95–2.67)) or CVD (1.23 (0.44–3.41)) death, high-fit/abdominal-obese men had no greater risk for all-cause (1.20 (0.80–1.78)) or CVD (0.62 (0.25–1.53)) death, and high-fit/percent body fat (%BF)-obese men had no greater risk for all-cause (1.19 (0.90–1.56)) or CVD (0.86 (0.52–1.43)) death compared with their high-fit/normal counterparts.

Conclusions

Fitness is a powerful effect modifier in the association of adiposity to mortality in men with HTN, negating the all-cause and CVD mortality risk associated with obesity.

Hypertension (HTN) is associated with increased all-cause and cardiovascular disease (CVD) mortality.¹ Currently, nearly 30% of the US adult population is hypertensive.² Higher levels of cardiorespiratory fitness (CRF)^3,4,5 and physical activity^6,7 can reduce the risk of HTN for healthy normotensive persons. Regular exercise can also lower resting blood pressure in hypertensive adults.⁸ However, the influence of fitness on all-cause and CVD mortality in people with HTN is not fully understood.

Currently, over half of those with HTN are obese⁹ and the risk of HTN ranges from two¹⁰ to five times¹¹ higher among obese individuals than among those of normal weight. Higher CRF reduces the mortality risks associated with both HTN^12,13 and obesity.^14,15,16 Few studies, however, have simultaneously examined the relation of fitness and adiposity to mortality in hypertensive persons and all of these studies assessed adiposity from body mass index (BMI).^12,13,15 Although BMI is well correlated with %BF,¹⁷ at a given BMI there is considerable variation. For example, in subjects with a BMI of 25 kg/m², body fat percentages can range from 20 to 50% (ref. 18). Therefore, other clinical measures of adiposity, such as %BF and waist circumference (WC), further elucidate associations among adiposity, fitness, and mortality in hypertensive persons.

A previous report from our group provided compelling evidence that moderate to high levels of fitness can reduce all-cause mortality risk in hypertensive men.¹² This report, however, did not examine the combined effects of fitness and adiposity on mortality. Such joint analyses may identify associations obscured in independent analyses alone. The purpose of this study was to examine the independent and joint effects of fitness and various clinical measures of adiposity (BMI, WC, and %BF) on all-cause and CVD mortality in men with HTN.

Methods

Study population. The Aerobics Center Longitudinal Study is an ongoing, prospective epidemiologic study of patients examined at the Cooper Clinic in Dallas, Texas. Study participants came to the clinic for periodic preventive health examinations and for counseling regarding diet, exercise, and other lifestyle factors associated with increased risk of morbidity and mortality. Many participants were referred by their employers; others were referred by their physicians or were self-referred. Between 1974 and 2003, 35,151 men received a comprehensive medical examination and were enrolled in the study. Most participants are Caucasian and from middle and upper socioeconomic strata. The current analysis included participants who at baseline were free of self-report history of myocardial infarction or stroke, and cancer; had normal resting electrocardiograms, had complete data on all three adiposity measures, and were able to complete an exercise test to at least 85% of their age-predicted maximal heart rate. We excluded men with a BMI <18.5 kg/m² at the baseline examination. All participants had HTN based on a self-report history of physician diagnosis or a measured resting systolic or diastolic blood pressure of ≥140 or ≥90 mm Hg, respectively, at the baseline examination. A total of 13,155 men with HTN between the ages of 20 and 84 years met our inclusion criteria and were evaluated. Written informed consent was obtained from participants before enrollment into the follow-up study, and the study was reviewed and approved annually by the Cooper Institute Institutional Review Board.

Clinical examination. A standardized baseline physician examination, including personal and family histories, and clinical assessment, including fasting blood chemistries and a maximal symptom-limited treadmill exercise test were conducted as described previously.^4,12 BMI (kg/m²) was computed from measured height and weight. Percent body fat (%BF) was assessed using hydrostatic weighing, the sum of seven skinfold measures, or both methods, following standardized protocols.¹⁹ Detailed description of our hydrodensitometry procedures have been published elsewhere.¹⁴ WC was measured at the level of the umbilicus. Adiposity exposure groups were based on standard clinical definitions for: BMI (normal weight 18.5–24.9, overweight 25.0–29.9, obese ≥30.0 kg/m²); WC (normal ≤102.0 cm; abdominal obesity >102.0 cm); and %BF (normal <25%; obese ≥25%) (ref. 20). Following a brief period of sitting quietly, resting blood pressure was measured in the seated position using auscultatory methods with a mercury sphygmomanometer. Systolic and diastolic pressures were recorded as the first and fourth Korotkoff sounds, respectively. Serum samples were analyzed for lipids and glucose using standardized bioassays. Baseline diabetes and hypercholesterolemia were defined as a history of physician diagnoses, measured phenotypes that met clinical thresholds for a specific condition, or when appropriate the combination of both methods. Smoking habits (never, former, and current smoker), alcohol, and physical activity habits (physically inactive or not) were obtained from a standardized questionnaire. Physical inactive was defined as reporting no leisure-time physical activity in the 3 months before the examination. Drinks per week of alcohol intake were computed with one drink standardized to 12 ounces (3.41 dl) of beer, 5 ounces (1.421 dl) of wine, or 1.5 ounces (0.4262 dl) of hard liquor.

CRF was quantified as the duration of a maximal treadmill exercise test using a modified Balke protocol.^4,21 The test was terminated when the men were exhausted or if the physician stopped the test for medical reasons. Total time of the test on this protocol correlates highly with measured maximal oxygen uptake in men (r = 0.92) (ref. 22). The percentage of age-predicted maximal heart rate (220-age) that was achieved during exercise testing was 101 ± 7%. To standardize exercise performance, we estimated maximal metabolic equivalents (1 metabolic equivalent = 3.5 ml VO₂ uptake/kg/min) from the final treadmill speed and grade (VO₂ (ml/kg/min) = (0.1·speed) + (1.8·speed·grade) + 3.5 ml/kg/min) (ref. 23). In previous Aerobics Center Longitudinal Study reports that have shown low CRF is an independent predictor of mortality,¹⁵ we have defined low, moderate, and high CRF exposures according to the lowest 20%, and the middle and the upper 40%, respectively, of the age- and sex-specific distribution of maximal exercise duration in the overall Aerobics Center Longitudinal Study population.²⁴ To maintain consistency in our study methods and because a widely accepted clinical categorization of CRF does not exist, we used the above approach.

Mortality surveillance. Vital status was ascertained using the National Death Index and using death certificates from states in which participant deaths occurred. Over 95% of mortality follow-up is complete by these methods. The National Death Index has been found to provide mortality data that are of similar accuracy to those determined by an Endpoints Review Committee, which reviews both death certificates and relevant medical records to make their determination. Sesso et al.²⁵ compared cause of death determined by an Endpoints Committee to those from National Death Index in the Physicians Health Study for deaths occurring between 1982 and 1998. For National Death Index, the sensitivity for CVD mortality was 90% and the specificity was 93%. Causes of death were identified using International Classification of Diseases, Ninth Revision (ICD-9) codes before 1999, and the Tenth Revision (ICD-10) codes (in parentheses) during 1999–2003: CVD, 390–449.9 (I00-I78).

Statistical analysis. Descriptive statistics summarized baseline characteristics by fitness levels. Trends in covariates over fitness levels were assessed using F tests. The follow-up interval was computed from the date of a participant's baseline examination until the date of death for decedents, or until 31 December 2003 for survivors. The mean ± s.d. follow-up interval was 12.2 ± 7.4 years. Cox regression analysis was used to estimate hazard ratios (HRs) and 95% confidence intervals of all-cause and CVD mortality according to fitness and adiposity exposure categories. The proportional hazards assumption was examined by comparing the cumulative hazard plots grouped on exposure; no appreciable violations were noted. All multivariable analyses included the following commonly used potential confounders: age (years), examination year, physical activity (inactive or not), smoking status (never, past, and current smoker), alcohol intake (≥5 drinks/week, yes or no), resting systolic and diastolic blood pressure (mm Hg), diabetes (present or not), hypercholesterolemia (present or not), and family history of CVD (present or not). Tests of linear trends in mortality rates and risk estimates were computed using ordinal scoring for fitness and adiposity groups. We estimated the population attributable risk of mortality for unfit, obesity, and other mortality risk factors to quantify their possible influence on all-cause and CVD mortality in our population sample. population attributable risk was computed as P_c(1 − 1/HR_adj), where P_c is the prevalence of a risk factor among decedents, and HR_adj is the multivariable adjusted HR for mortality associated with the specified risk factor.²⁶ Finally, we tested joint associations of adiposity and fitness with all-cause and CVD mortality. We separately entered and tested each of the crossproducts between fitness and adiposity in the Cox regression models and found no significant interactions among exposure groups. All P values are two-sided and P < 0.05 was regarded as statistically significant.

Results

There were 883 deaths (355 due to CVD) during 160,886 man-years of exposure. Overall, the mean age (±s.d.) of participants was 47.7 (±9.9) years. Compared with higher fit individuals (Table 1), those in the low-fitness category were younger, had higher fatness, were more likely to be current smokers, and CVD risk factors were less favorable.

Table 1

Baseline characteristics according to cardiorespiratory fitness category, Aerobics Center Longitudinal Study, 1974–2003

Rates and HRs of all-cause and CVD mortality according to CRF groups among men with HTN are shown in Table 2. An inverse gradient of overall death rates was observed across incremental CRF groups (P_trend < 0.0001). After adjusting for age, examination year, smoking, drinking, resting blood pressure, chronic conditions, and family history of CVD, men with moderate and higher CRF had 45 and 58% lower risks of total mortality than men with low CRF (P_trend < 0.0001). The inverse association between CRF and all-cause mortality remained significant after additional adjustment for BMI, WC, and %BF (P_trend < 0.0001). Similar inverse patterns of association were observed between CRF and CVD death (P < 0.0001).

Table 2

Rates and hazard ratios for mortality, according to baseline cardiorespiratory fitness levels

Next we evaluated mortality across adiposity categories (Table 3). When adjusted for the same set of covariables as in the CRF analyses, those with BMI-defined obesity, abdominal obesity, and high %BF had a higher overall mortality risk compared with their peers in normal weight or adiposity groups, respectively (all P < 0.05). The positive relationships persisted for BMI and WC, after further adjustment for fitness (P < 0.05 for each), however, the positive relationship for %BF was no longer significant. Adjusted models showed similar patterns of association between BMI, WC, and %BF with CVD death. However, additional controlling for fitness attenuated the association between BMI and CVD death (HRs for overweight and obese: 1.00 and 1.45, P_trend = 0.046). No associations were noted between WC and %BF and CVD death after accounting for fitness.

Table 3

Rates and hazard ratios for mortality, according to baseline adiposity measures

Table 4 shows the relative importance of unfit, obesity, and other mortality risk factors in terms of population risks. After adjustment for age, examination year, resting blood pressure, and each risk factor in the table, unfit and BMI-defined obesity were significantly associated with all-cause and CVD mortality. However, no association was observed among abdominal obesity, abnormal %BF, and mortality in men with HTN. To place the risk of mortality for each exposure in the context of population-disease burden, we estimated population attributable risk based on the baseline prevalence and strength of association with mortality for each exposure. Therefore, if all hypertensive individuals in our study group were fit, we might expect that there would have been 11% fewer all-cause deaths and 16% fewer CVD deaths. However, if all hypertensive individuals were not obese (BMI <30), we might only expect a 5% fewer all-causes deaths and 8% fewer CVD deaths.

Table 4

Hazard ratios and population attributable risk of mortality for unfit, obesity, and selected mortality risk factors among men with hypertension

Finally, results of the joint effects of fitness and fatness on all-cause and CVD mortality are presented in Table 5. Compared with high-fit/normal-weight men, low-fit/obese men had 3.0- and 4.4-fold higher risks of overall and CVD death, respectively. Compared with high-fit/normal WC men, low-fit/abdominal obese men had 2.6- and 3.7-fold higher risks of overall and CVD death, respectively. Compared with high-fit/normal %BF men, low-fit/high %BF men had a 2.4- and 3.4-fold higher risk of overall and CVD death, respectively.

Table 5

Joint associations of cardiorespiratory fitness and adiposity exposures with all-cause and CVD mortality among men with hypertension

Discussion

The objective of this study was to evaluate associations among CRF, adiposity, and mortality in men with HTN. We found that higher levels of CRF are protective against all-cause and CVD mortality in this population. In the fully adjusted model, moderate-fit and high-fit men with HTN as compared with low-fit men with HTN were 42 and 57% less likely to die for any reason, respectively (Table 2). Our findings compare favorably to the Lipids Research Clinics Prevalence Study of 1,009 hypertensive men in which those in the highest fitness quintile had approximately half the all-cause and CVD mortality risk compared to men in the lowest quintile.¹³ Together these results underscore the importance of fitness for reducing mortality risk in populations at higher risk for CVD because of elevated blood pressure.

Data on the joint effects of adiposity and CRF on mortality are sparse, as objective measures of fitness (maximal exercise testing on a treadmill) are not widely available, nor are data on %BF or WC. Stratifying adiposity by CRF permits more rigorous analyses of combined subgroups than adjustment for either variable alone. Yet to our knowledge, no study has simultaneously examined CRF and different measures of adiposity specifically in subjects with HTN. In this study, we observed a strong inverse relation between fitness and mortality such that hypertensive men with normal BMI, WC or %BF were more likely to survive only if they possessed high fitness (Table 5). Our findings on BMI are in agreement with our previous studies of the broader Aerobics Center Longitudinal Study cohort^14,15 and compare favorably with a recent report from the Kuopio Ischemic Heart Disease Risk Factor Study,¹⁶ which found that CRF reduced all-cause mortality risk by 57% in patients having a combination of two or three risk factors, which included HTN, BMI ≥25 kg/m², and smoking.

There are several possible explanations for the lack of association of abdominal obesity and %BF to CVD mortality in this study. First, although WC and %BF are standard clinical measures of adiposity, we did not have data on extremity circumferences. Greater lower body circumferences at a given BMI or WC level appear to be cardioprotective.²⁷ Second, our finding that abdominal obesity increased all-cause mortality risk by 20% is nearly identical to that reported in a recent large scale life insurance study of men,²⁸ but data on CVD mortality are sparse.^29,30 Third, we are unaware of any previous study that specifically examined the relation of %BF to CVD mortality. To understand the independent associations of WC and %BF to CVD mortality will require further study.

Our main finding that higher CRF reduced the risk for all-cause and CVD mortality is consistent with several studies reporting significant training-mediated reductions in systolic and diastolic blood pressure in adults with HTN.⁸ Neurohormonal adaptations to exercise that may contribute to this training-induced hypotensive effect include reductions in plasma norepinephrine, plasma renin, and angiotensin.⁸ Other mechanisms include improved vascular responsiveness³¹ and structural changes in skeletal muscle.³² These well-documented physiological responses to exercise training underscore the importance of regular exercise for reducing mortality risk in individuals with HTN, as well as preventing the development of HTN in normotensive persons.

We computed population attributable risk values to estimate the burden of all-cause and CVD mortality attributable to low fitness, obesity, and other risk predictors. If all unfit hypertensive men in our population sample became fit, 11 and 16% of the deaths might have been averted from all-cause and CVD, respectively. Currently, there are not enough data to determine how much of the mortality burden in population with HTN may be due to unfit or/and obesity.

Our study has several strengths. We measured cardiovascular fitness levels objectively by maximal exercise testing, and included data on %BF and WC in addition to BMI measures of adiposity. All subjects underwent an extensive physical examination, which provides thorough information on the presence or absence of baseline disease. The mean follow-up of >12 years for >13,000 men with HTN enabled us to assess joint associations of CRF and three clinical measures of adiposity to all-cause and CVD mortality. To our knowledge, there is no previous epidemiologic study of these associations in hypertensive persons.

Our study also has several limitations. First, although we utilized standard clinical measures of adiposity, these methods are not optimal. BMI and WC are only proxy measurements of body fatness and %BF was estimated from skinfold measures or hydrostatic weighing each of which have well-known methodological limitations.²³ Second, our study included primarily white men from middle to upper socioeconomic status. Therefore, the results may not apply to women or other groups of men. Third, CRF is a single measure that is influenced by many factors, including age, heredity, and recent and lifelong activity patterns.²³ The extent to which fitness may be improved in adults, or the influence this may have on mortality, cannot be determined from the present investigation. Fourth, as we only have baseline data on weight, CRF and other exposures, we do not know whether changes in any of these variables occurred during follow-up and how this might have influenced the results. It is possible that many men with HTN were treated at some point in the follow-up interval. Others may have experienced increases/decreases in the above exposures. Such misclassification of exposure usually causes an underestimate of the association, but it is possible that it might overestimate the association. Fifth, we were unable to evaluate hypertensive medication use, including specifics on types of antihypertensive agents, dosages, duration of treatment, or efficacy of treatment. However, it seems unlikely that these would be significant confounders in this well-educated population of middle to upper socioeconomic strata men who had access to excellent medical care. Sixth, this study did not have information on baseline serum creatinine or estimated glomerular filtration rate to assess the impact of chronic kidney disease on the results. Finally, as blood pressure was measured during a single visit, misclassifications in our diagnosis of HTN may have occurred.

In conclusion, this prospective epidemiologic study of men with HTN found that higher levels of CRF were associated with lower all-cause and CVD mortality risk regardless of obesity status defined by BMI, WC, or %BF. Hypertensive men of normal-weight survived better only if they registered high fitness. Importantly, the incremental reduction in CVD mortality risk was greater from low to moderate (~50%), than moderate to high fitness (~20%) supporting the premise that it may be possible to achieve a sizable reduction in CVD mortality simply by encouraging hypertensive individuals of low fitness, including those who are obese, to achieve moderate fitness. Most individuals can prevent falling into the low CRF category by engaging in regular physical activity, such as brisk walking for 30 min ≥5 days per week.³³ Also, the favorable effects of regular exercise on blood pressure and improved fitness in patients with HTN may in turn reduce hypertensive medication use along with associated health-care costs.

Disclosure

S.N.B. receives book royalties (<$5,000/year) from Human Kinetics; honoraria for service on the Medical Advisory Boards for Matria Health Care and Jenny Craig; and honoraria for lectures from scientific, educational, and lay groups. He gives these fees to the University of South Carolina Educational Foundation or to other nonprofit groups. During the past 2-year period, he has received a research grant from BodyMedia, Inc. Other authors declared no conflict of interest.

Acknowlegements

We thank the Cooper Clinic physicians and technicians for collecting the baseline data, and staff at the Cooper Institute for data entry and data management. This work was supported by National Institutes of Health grants AG06945 and HL62508.

© American Journal of Hypertension, Ltd. 2009