Can Weight Control and Regular Physical Activity Increase Survival in CHD Patients?^∗

Introduction

The Centers for Disease Control and Prevention reports that coronary heart disease (CHD) is responsible for >370,000 deaths annually, or about 50% of all heart disease related deaths (1). About 735,000 Americans have a heart attack each year, and for 525,000 of them, it is a first heart attack. Thus, >200,000 heart attacks occur annually in people who had already experienced a first episode. Surviving a second CHD event is therefore an important secondary prevention goal.

Recommendations for enhancing survival following a CHD episode include no smoking, a healthy diet, body weight control, and regular physical activity (PA), among others (2). Of particular interest here are weight control and PA level. Both are known to be risk factors for incident CHD (3), but they also relate to survival in CHD. For instance, among patients with CHD, high variability in body weight over a 4.7-year period was associated with higher mortality (4). Cardiac rehabilitation programs with a PA component have been shown, in a meta-analysis of 34 studies, to reduce all-cause and CHD mortality (5). Notably, being physically active was associated with reduced all-cause and cardiovascular mortality in a 12-year-follow-up period in 6,493 patients with CHD of the HUNT (Nord-Trøndelag Health) study whereas, in a paradoxical manner, overweight and obese patients also had reduced mortality (6). This so-called obesity paradox has been found in multiple studies and confirmed in a meta-analysis of 40 cohort studies comprising >250,000 patients with CHD in which overweight and obese CHD patients had a lower risk of all-cause and cardiovascular disease (CVD) mortality compared with normal-weight CHD patients (7). However, there is a paucity of data on the effects of long-term changes in body weight and PA level on mortality in individuals with CHD.

In this issue of the Journal, Moholdt et al. (8) report on a secondary analysis of the HUNT Study based on 3,307 men and women with CHD in which 1,493 deaths (55% from CVD) were recorded over 30 years of follow-up. Patients were examined at 3 time points and the median duration of the follow-up reached 15.7 years. CHD patients with data on PA, body mass index (BMI), diabetes mellitus, self-reported health, blood pressure, smoking, and alcohol consumption in 2 or 3 HUNT study waves were included. Height and weight were measured and changes in body mass over time were reported in kg/m²/year. PA level was estimated at each examination by questionnaire probing frequency, duration, and intensity of leisure time PA (9).

The main findings are that although weight loss was associated with higher all-cause mortality in those who were normal weight at baseline, weight gain was not associated with increased mortality. All-cause mortality was lower in participants who maintained a low or a high PA level compared with those who remained inactive over time. Similar associations were observed for CVD mortality. This study contributes important new information on the central question of whether long-term changes in body weight and PA level influence mortality rates after a CHD event. However, there are a number of issues that could not be addressed by Moholdt et al. (8), and other considerations that need to be taken into account when interpreting their findings.

One of these issues relates to the measurement of BMI and the assessment of PA levels. In the present study, height and weight were directly measured at each of the 3 examinations. The authors opted to use BMI changes instead of body weight fluctuations. It remains unclear which of change in body weight or change in weight relative to height is the most useful indicator of body mass fluctuations in the context of health outcome prediction. However, the reliance on BMI does not simplify the translation of findings into practical guidelines. For instance, it is not readily clear how much body weight loss is involved in a category defined as a loss of >0.10 kg/m²/year. From the paper, the mean baseline height was 1.68 m and weight 74.9 kg, for a mean BMI of about 26 kg/m². The cutoffs used translate into a weight gain or loss of ≥9 kg over 30 years. Such large body weight excursions are likely accompanied by higher total adiposity, visceral fat deposition, and ectopic fat accumulation in liver and potentially other organs (10,11). These undesirable traits are entrained by weight gain in a large fraction of adults, but not in all, which implies that not all weight gain is atherogenic, as suggested by the obesity paradox.

In the present study, PA level was assessed by questionnaire. Because of changes in the questions used in the second examination compared with the other examinations (9), it is not clear if the PA level classification was affected. Participants were grouped into 3 PA categories at each examination, namely inactive, low PA, and high PA. The high-PA group includes those who met the recommended PA guidelines plus those who exceeded them. The low-PA group is composed of those who did not meet the PA guidelines, a category that is likely to overlap with the distribution of those classified as inactive. Despite these limitations and the “softness” of the questionnaire-based assessment, a strong relation emerged between PA level changes over time and mortality rates in CHD patients, with the consistently active experiencing the lowest death rates from all causes and from CVD. It is important to recognize that assessment of PA level by self-reports results in much higher rates of adults being classified as active than when PA is measured by accelerometer or pedometer (12–14). For instance, in the National Health and Nutrition Examination Survey adult population, only 5% met the recommended PA level when it was assessed by accelerometer (14), a percentage almost 10-fold below the commonly reported questionnaire-based level.

When evaluating the effects of weight and PA level changes over time on any outcome, it is important to remember that changes in PA are correlated with the changes in weight. For instance, in the same HUNT study cohort, it was shown in an earlier publication that men or women maintaining a PA level above the recommended guidelines gained only about one-half as much weight as did those who remained inactive over time (9). In the present study, weight and PA changes were reciprocally adjusted for (in model 3), which has largely attenuated the confounding effect resulting from their lack of independence.

In summary, the study of Moholdt et al. (8) adds to our understanding of the role of long-term changes in body weight and PA level on mortality rates of individuals who have already experienced a CHD event. They could not document an increase in mortality among CHD patients who gained weight during the follow-up period, while weight loss was associated with higher all-cause mortality in those who were normal weight at baseline. In contrast, maintaining or increasing PA level at or above the recommended guidelines resulted in lower all-cause and CVD mortality rates. The findings add to the body of data suggesting that promoting regular PA in CHD patients is likely to save lives.

Footnotes