From the HealthPartners Research Foundation, Minneapolis, MN.
Correspondence: Kenneth F. Adams, 21111R, HealthPartners Research Foundation, 8170 33rd Ave S., P.O. Box 1524, Minneapolis, MN 55440–1524. E-mail: email@example.com.
The majority of Americans enter midlife overweight or obese. With this excess weight come other cardiometabolic risk factors, and an increasing proportion of adults are living with chronic disease. For the growing segment of the middle-aged and older population who are overweight or mildly obese, the question might reasonably be asked, “Does losing weight improve longevity?” Conversely, “Is it harmful for persons who are already overweight to continue gaining weight into old age?”
In this issue of Epidemiology, Myrskylä and Chang1 report that large weight loss was related to elevated mortality, with a much stronger association in persons with low body mass index (BMI) at baseline. Large weight gain was associated with mortality, but only in persons who were initially obese. No data were available to indicate whether subjects intended to lose weight. Although the authors do not state explicitly that they were evaluating the effects of purposeful or intentional weight loss, their extensive efforts to exclude the influence of disease-related weight loss in their analyses suggest that this was their aim. They concluded that they found no benefits to mortality from weight loss, whereas weight gain seemed to be harmless except in very obese persons.
Can observational studies, using only observed body weight change and limited behavioral and health data, answer questions about the long-term benefits and harms of weight loss? Many previous studies have reported elevated risk of death in persons who have lost weight. Problems with interpreting these studies have been well described.2–7 The fundamental issue is that unintentional weight loss is common in older people8,9 and often caused by underlying disease or psychosocial problems that are themselves associated with increased mortality. If these conditions are undetected (or if necessary data are not available to the researcher), it is difficult to account for their effects. More recent studies have asked subjects whether they intended to lose weight. This strategy represents an advance, but it only partially disentangles intentional from unintentional weight loss7: among persons expressing the intent to lose weight, a substantial proportion of actual weight loss is likely the result of other causes such as stress or disease.7,10
To move the field forward, it will be necessary to collect much richer, denser, and more biologically relevant data. Others have already described the data and design elements needed to advance our understanding of the long-term harms and benefits of weight change.3,6,7,11 I will highlight 3 of these elements.
In a commentary in Epidemiology 10 years ago, Tim Byers stated that epidemiologists had reached the limits of what could be learned in observational studies that simply estimate associations between undifferentiated weight loss and mortality.11 Byers called on epidemiologists to engage a major effort to uncover the specific reasons for weight change. Similar comments have been made by others.3 Byers noted that episodes of substantial weight loss, including unintentional weight loss, are common but poorly understood. He suggested the possibility of unintentional weight loss syndromes that could be precursors to chronic disease. Researchers now recognize and are beginning to understand several weight loss syndromes in older persons and in those with disease.12–14 It is possible that these or similar syndromes occur in younger, apparently healthy persons, and also affect functional status and mortality. Byers also noted that intentional weight loss could have long-term health consequences in persons employing nutritionally unsound practices. All of these possibilities require more intensive data collection for full evaluation.
A second need is consideration of weight pattern over the lifetime.3,11,15 Most mortality studies consider weight loss relatively late in life, when disease may have a disproportionate influence. Influence of body mass or weight gain in young adulthood should also be taken into account. Body mass considered at younger age is more strongly related to mortality than body mass later in life.16 The effects of adiposity are likely cumulative over time and have a latent period. Excessive weight gain in adolescence or young adulthood may lead to long-term adiposity, which could be more harmful than midlife weight gain. Alternatively, the weaker associations in older persons could be explained by weight loss at midlife due to obesity-related disease.17 Truly longitudinal studies, making use of repeated measurements, may bring a better understanding of temporal relationships such as those among depression, weight change, and physical disability.18
A third way to move this research forward is through data on body composition, metabolism, and other biologic measures.3,19 Weight loss leads to loss of bone, lean mass (muscle and organ) and fat mass (adipose). Allison et al19 demonstrated that loss of fat mass is associated with decreased mortality, while suggesting that loss of lean mass increased mortality risk. Research at this physiologic level of detail is important both because it furthers biologic understanding of the weight loss process and because it suggests possible interventions for healthy weight loss. This approach has been implemented best in aging research. Aging typically leads to loss of lean body mass and muscle strength, whereas fat mass is increased, particularly intra-abdominal visceral fat. Weight tends to remain stable or decline, masking increased adiposity. Several weight loss syndromes such as sarcopenia (loss of muscle)13 and inadequate caloric intake12 are common enough to be considered normal processes, whereas cachexia14 and wasting are the result of disease. The Health ABC Study is one study specifically designed to evaluate the influence of changes in body composition on outcomes such as death and functional disability in the elderly. The study is collecting the relevant physiologic, metabolic, biochemical, and behavioral and outcome data necessary to understand the complexity of these relationships.
To advance our understanding of the benefits or harms of weight change, it is important that future research identifies the specific reasons for weight change, the physiologic mechanisms involved, and the populations (eg, age groups) to whom results apply. Previous commentators have identified the improved data and study designs needed to answer these questions. Randomized trials can address many questions, but observational studies are still necessary to inform these trials and to address issues not conducive to interventions.
ABOUT THE AUTHOR
KENNETH ADAMS is a research investigator at HealthPartners Research Foundation in Minneapolis, MN. He has published detailed analyses relating, body mass to mortality, colorectal cancer, and renal cell cancer. He is interested in creating virtual cohorts from electronic medical record data, to examine relations among body mass, clinical measures, and obesity-related disease.
1. Myrskylä M, Chang VW. Weight change, initial BMI, and mortality among middle- and older-aged adults. Epidemiology
2. Willett WC, Stampfer M, Manson J, VanItallie T. New weight guidelines for Americans: justified or injudicious? Am J Clin Nutr
3. Kuller L, Wing R. Weight loss and mortality. Ann Intern Med
. 1993;119(7 pt 1):630–632.
4. Williamson DF, Pamuk ER. The association between weight loss and increased longevity. A review of the evidence. Ann Intern Med
. 1993;119(7 pt 2):731–736.
5. Lee IM, Paffenbarger RS Jr. Is weight loss hazardous? Nutr Rev
. 1996;54(4 pt 2):S116–S124.
6. Fontaine KR, Allison DB. Does intentional weight loss affect mortality rate? Eat Behav
7. Yang D, Fontaine KR, Wang C, Allison DB. Weight loss causes increased mortality: cons. Obes Rev
8. Wallace JI, Schwartz RS, LaCroix AZ, Uhlmann RF, Pearlman RA. Involuntary weight loss in older outpatients: incidence and clinical significance. J Am Geriatr Soc
9. Sahyoun NR, Serdula MK, Galuska DA, Zhang XL, Pamuk ER. The epidemiology of recent involuntary weight loss in the United States population. J Nutr Health Aging
10. Allison DB, Heo M, Fontaine KR, Hoofman DJ. Body weight, body composition, and longevity. International Textbook of Obesity
. Sussex, United Kingdom: Wiley; 2001.
11. Byers T. The observational epidemiology of changing weight: an appeal for reasons. Epidemiology
12. Morley JE. Decreased food intake with aging. J Gerontol A Biol Sci Med Sci
. 2001;56(Spec No 2):81–88.
13. Roubenoff R. Sarcopenia: effects on body composition and function. J Gerontol A Biol Sci Med Sci
14. Morley JE, Thomas DR, Wilson MM. Cachexia: pathophysiology and clinical relevance. Am J Clin Nutr
15. Manson JE, Stampfer MJ, Hennekens CH, Willett WC. Body weight and longevity. A reassessment. JAMA
16. Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL. The effect of age on the association between body-mass index and mortality. N Engl J Med
17. Losonczy KG, Harris TB, Cornoni-Huntley J, et al. Does weight loss from middle age to old age explain the inverse weight mortality relation in old age? Am J Epidemiol
18. Forman-Hoffman VL, Richardson KK, Yankey JW, Hillis SL, Wallace RB, Wolinsky FD. Impact of functional limitations and medical comorbidity on subsequent weight changes and increased depressive symptoms in older adults. J Aging Health
19. Allison DB, Zannolli R, Faith MS, et al. Weight loss increases and fat loss decreases all-cause mortality rate: results from two independent cohort studies. Int J Obes Relat Metab Disord