A Research Agenda: The Changing Relationship Between Body Weight and Health in Aging

It is increasingly recognized that changes in weight and body composition with age are strongly connected with health status and physical function. Aging is typically associated with reductions in total and lean mass, so that in the last few years of life, older people frequently report having lost weight and strength (1–3). We know only a few important facts about this process. On average, muscle mass declines with age, and even in older persons with stable weight, muscle is replaced by fat over time (4,5). Increasing fatty infiltration of muscle tissue is associated with decreasing muscle strength (6). The general pattern of weight change over the lifetime is that weight increases through approximately age 60 years and decreases thereafter (7). Weight gain from early adulthood through midlife is related to increases in both fat and muscle mass, but weight loss at older ages is associated with higher risk for a disproportionate decline in muscle mass (1,8).

In addition to changes in fat and muscle mass, fat location and muscle quality change with age. Waist circumference and intraabdominal visceral fat increase with age at a greater rate than total weight, reflecting changes in the distribution of fat (9,10). The extent to which individual trajectories of weight and body composition mirror these population average trajectories is unclear (11). Case studies and clinical experience suggest that most individuals actually maintain weight until a period of accelerated body composition change occurs, which parallels deterioration of health status. However, the critical age of acceleration is highly variable between individuals and is perhaps a marker of underlying biological change, as opposed to chronological age.

Age-related changes in body composition have important implications for health in late life. Obesity is clearly associated with disability, but the mechanism for this association is not clear. In studies that obtained detailed measures of body composition, muscle strength and fat mass independently predict incident disability and mobility limitation, while there is less evidence for an association between muscle mass and disability (10,12–17). It is unclear whether obesity and muscle mass or strength act as synergistic risk factors in producing excess risk of disability (14,15,17–20). While guidelines exist for healthy body mass index (BMI) with age, consideration of body composition may also be important in clinical assessment.

Recently, a working group met in Erice, Italy, to discuss gaps in our understanding of causes and consequences of changes in body weight and composition over the aging process. During a number of formal presentations and in informal discussions that followed, we identified critical research questions that would advance our understanding of the effect of changes in weight and body composition on health, function, and quality of life in older persons. In an effort to focus research on potentially high impact areas, we have selected three of these questions as immediate priorities for publication in an upcoming special section of JGMS.

1. What factors underlie the changing relationships between BMI and health with increasing age? How can we dissect BMI and health relationships to understand the health effects of weight on disease, disability, and death?

Clearly, BMI is importantly related to chronic conditions, disability, and mortality. However, for many of these outcomes, relationships between BMI and health change with age (7,21). For instance, the relationship between BMI and mortality is increasingly U shaped with advancing age, with higher mortality among both underweight and obese older persons, and the BMI associated with the lowest mortality increases with age (22,23). BMI is a less reliable indicator of fatness in older people because of changes with body composition in aging, including height loss and increases in body fat even among weight-stable older persons. BMI also does not take fat distribution into account, which is known to influence health (24,25). Additionally, if recent weight loss is not taken into account, BMI may be an indicator of underlying health status, reflecting a reverse association in which disease influences BMI, in contrast to the typically theorized relationship in which BMI predicts health status (26–28).

Both the factors that predict BMI and its role in health outcomes change with age, but the causes of this change are unclear. We would like to encourage papers that identify key factors needed to evaluate associations between body mass index and health in old age. For instance, what measures of body composition are most predictive of health outcomes for people with different BMIs? These might include anthropometric and strength measures, measures of body composition, biomarkers, or other health indicators. Predictors for important health outcomes may differ in those with extreme values of traditional risk factors (29). Given greater variability in health outcomes within BMI categories at older ages, what characteristics distinguish healthy and frail older persons at different points on the BMI distribution? Are these characteristics different at both extremes of body mass—underweight and obese? To what extent are associations between BMI and health in old age driven by body composition, behavior, and/or underlying health status? How and why do the effects of BMI depend on overall health? Are there relevant animal models that would elucidate mechanisms through which BMI and body composition affect health at older ages?

2. What are the lifetime characteristics of those who are obese and functionally impaired in old age?

On average, obesity is associated with higher levels of functional impairment and disability (23,30–33). However, there is significant heterogeneity in the relationships between weight, muscle mass, strength, and function. Given the rising prevalence of obesity, the most common phenotype of frailty in the future may be an obese, disabled older person. It is important to understand why sarcopenia, functional impairment, frailty, and disability occur in some obese persons, but not in others. What factors measured earlier in adulthood might distinguish between those who are overweight and aging healthfully and those most likely to suffer the health consequences of overweight? How do the predictors of poor health or functioning vary in obese and nonobese populations?

3. Is weight loss in old age ever beneficial?

In what circumstances?

Numerous observational studies have reported associations between weight loss at older ages and higher mortality (26–28,34). Although few of these studies have been able to identify whether weight loss was intentional (35), this evidence, combined with results showing a higher ideal BMI at older ages, has led some to conclude that weight loss of any kind is inadvisable for older people (36). At the same time, short-term intervention trials have demonstrated that intentional weight loss is associated with declines in cardiovascular risk factors (37) and improvements in physical function (38,39) and strength (40) among overweight older persons, and intentional weight loss is associated with decreased mortality in animal models (41). American Society for Nutrition and NAASO Obesity Society guidelines recommend weight loss therapy that minimizes muscle and bone loss “for older persons who are obese and who have functional impairments or medical complications that can benefit from weight loss” (21). It is critical that we develop a knowledge base addressing the efficacy and safety of weight loss in older persons. Is there a safe and healthy way to lose weight in old age? How do we distinguish older persons who might benefit from weight loss interventions from those for whom weight loss would be harmful? What is the state-of-the-art knowledge about what interventions help to maintain bone and lean mass with weight change?