Weight loss is quite prevalent in the United States, with an estimated 40% of women and 25% of men trying to lose weight at any given time (118,119,132,154). Efforts to lose weight are also quite prevalent among teenagers, especially girls (118). Despite spending in excess of $30 billion annually to lose weight (132), weight gain is also quite common (151), and Americans are heavier now than ever before (66). Chronic weight fluctuation may be the norm for many Americans (91,132). Although most people lose weight to improve appearance, concern for health is frequently cited ac a reason to lose weight (132). It is for health reasons that weight loss is recommended for overweight individuals (91,92). The rationale for this stems from the prevailing consensus that obesity is a major cause of premature death (2,79,92) and that weight loss can be expected to improve the health and longevity prospects of overweight people (2,79,91). However, whether it is best to be thin, at least in terms of longevity, and whether weight loss improves health and therefore reduces risk of premature death, is far from clear cut (3-5,14,34,38,40,55,57,121,137). Indeed, Americans' obsession with weight is not without risks, both physical and psychological (34,38,105,115,156). The purpose of this review is to examine the hypotheses that 1) lowest mortality rates are observed in thin men and women and 2) weight loss reduces mortality rates.
BODY WEIGHT AND MORTALITY
The concept of an optimal body weight for lowest mortality is largely attributable to the life insurance industry, more specifically the Metropolitan Life Insurance Company (MetLife), which introduced the terms "ideal body weight" in the early 1940s (86,87) and "desirable body weight" in 1959 (88). Although MetLife urged in 1983 that these terms no longer be used (89), the company nevertheless asserted that its tables indicated "the weights at which mortality is lowest-or longevity highest." This assertion, however, is not well supported by MetLife's actuarial source for its tables, the 1979 Build Study (22). For example, in men and women ages 40 and older it is not uncommon to find weights up to 40 pounds in excess of the upper limit of the recommended range that are associated with mortality rates not materially different from, and in some cases lower than, mortality rates within the recommended range (22). The reason for this inaccuracy in the tables results from the fact that MetLife did not adjust weight recommendations for age, and it is apparent from the 1979 Build Study that optimal weights for lowest mortality increased with age (4,22). This has been demonstrated in several epidemiological studies as well (3,23,29,78,113,114,125,131).
The body weight-mortality issue remains controversial. Some studies show a direct relationship between body mass index (BMI, kg·m−2) and mortality (51,69,81,150) and suggest that the optimal BMI for lowest mortality is well below that of the average U.S. adult. Others show a "J"- or "U"-shaped relationship, with higher mortality at the extremes of the BMI distribution (27,31,42,43,53,74,113,114,131,137,138,140,142,147,148). These studies indicate an elevated risk for thin as well as stout individuals. Even among studies that indicate higher risk at high BMI, the threshold at which risk increases significantly is not distinct and may be well above average BMI (31,39,113,114,127,131,137,147). A meta-analysis of the BMI-mortality literature in women, for example, indicated that BMI had little impact on all-cause mortality until BMI exceeded ∼35 (137). By contrast, a number of studies show no relationship between BMI and mortality (20,23,37,56,85,117,126,147,148), and some even indicate an inverse relationship between BMI and mortality (39,48,78). Collectively, the majority of epidemiological studies of the relationship between body weight and mortality do not support the contention that greatest longevity is enjoyed by individuals with lower-than-average BMI.
The discrepancies among studies is not easily reconcilable but undoubtedly reflects to some extent the inexactness of epidemiological research. Because epidemiological studies are observational in nature and rely upon circumstantial evidence, specific criteria must be met before any definitive conclusions can be drawn (94,124). Inconsistencies in the literature cannot be ignored, for they may necessitate revision, or rejection, of hypotheses (124). In this regard, the hypothesis that thinness confers longevity fails to satisfy the tenets of epidemiology (94) on several counts, most notably the failure to show consistency and repeatability among the various studies. Furthermore, it is difficult, if not impossible, to show absolute independence of BMI as a predictor of mortality rates because so many other factors covary with BMI and impact health status and mortality risk as well. Thus while obesity has been reported to be an independent risk factor for cardiovascular disease mortality (51), obesity is only "independent" of the other risk factors measured in the study but not necessarily independent of one or more of numerous other possible contributing factors not measured in the study. For heart disease alone, at least 246 potential risk factors have been identified (50), although certainly they are not all of equal importance. Nevertheless, it is doubtful that any epidemiological study could satisfactorily control for all, or even most, of them.
A high BMI, for example, rather than being a genuine contributing factor to premature death, could be a symptom of a sedentary lifestyle and low fitness level, both of which could be the true risk factors for early mortality. Men and women with high BMI (e.g., >27) are more likely to be physically inactive and unfit than are men and women with BMI <27 (120). Because physical inactivity and low fitness levels are associated with increased mortality rates (11,17,94,95,99), it is possible that the link between high mortality rates and high BMI reported in many studies results in large part from the greater prevalence of sedentary lifestyle and low aerobic fitness in the high BMI strata. Data from the Aerobics Center Longitudinal Study (11,17), for example, indicate that the lowest death rates are observed in men and women with the highest fitness levels, regardless of BMI. During an 8.5-yr follow-up, unfit men with BMI < 27 had a death rate 2.9 times that of men with BMI > 30 who were classified as having at least moderate-to-high levels of fitness (11). Data on Harvard Alumni indicate a similar result for BMI and physical activity, i.e., physically active men with high BMI had lower all-cause mortality rates than inactive men with low BMI (95). These data suggest that physical fitness and activity are more important than BMI as predictors of all-cause death rates.
In this regard, it is worth examining the frequently published assertion that obesity is the second leading cause of preventable death in the United States, accounting for approximately 300,000 deaths annually. This statistic has been cited by the American Dietetic Association (2), by the National Task Force on the Prevention and Treatment of Obesity (92), and by physicians arguing in favor of pharmacotherapy for obesity treatment (79). In each instance, the same source was cited for this statistic (84). However, McGinnis and Foege (84) made no such assertion about obesity as a cause of death. The 300,000 deaths per year were attributed to "diet/activity patterns," not obesity. It is unjustified to equate a physical trait with behaviors. While poor diet and a sedentary lifestyle may lead to obesity, the studies used to generate the statistic included men and women across all BMI strata, not just the obese. Lack of regular physical activity alone has been estimated to contribute to 250,000 deaths per year in the United States (99). Thus, continued use of this statistic (84) to allege that obesity is second only to tobacco use as a cause of preventable death in America is unwarranted.
POSSIBLE BENEFITS OF MODERATE OBESITY
The fact that some studies show no relationship between BMI and all-cause mortality in some cohorts (20,23,37,56,85,117,127,147,148), and in some instances an inverse relationship between BMI and mortality (39,48,78), suggests that there may be some counterbalancing influences of BMI on health status. A high BMI and/or body fat may increase risk for certain diseases yet provide some protection against others (93). Risk for some cancers may be lower in men and women with high BMI (10,39,52,54,59,61,128,136,141). This is especially true for lung cancer and is independent of smoking status (54,61). In a review of the literature, Kabat and Winder (54) concluded that "the overall consistency in the inverse association of body mass index with lung cancer in reported studies is impressive. It is noteworthy that no study shows a significant contrary trend." In general, diseases of the lung seem to be less common in fat men and women (34,54,113).
Osteoporosis, which afflicts > 25 million U.S. men and women over the age of 45 yr, is another disease for which a high BMI seems to be protective (32,110,116,135,143). The prevalence of this disease worldwide also suggests that osteoporosis is a growing international health care problem (9,143). An estimated 1.5 million osteoporosis-related bone fractures occur each year, resulting in significant morbidity and mortality (9,143). More women may die as a result of complications from osteoporosis-related fractures than from cancers of the cervix, uterus, and breast combined (9). Women with BMI > 26-28 generally have ∼4-25% greater bone density throughout their body compared with women with BMI <22-24 (143). This coincides with lower risk of fracture (143). While this should not be taken as a call for thin women to gain weight, it does suggest that moderate degrees of overweight/obesity are preferable to thinness in terms of minimizing risk of bone loss and related fractures.
Although many studies reveal a positive relationship between high BMI and heart disease mortality (e.g., 52,69,81), this is by no means a universal finding. The lack of a relationship between BMI and coronary heart disease (CHD) mortality reported in several cohorts with follow-up ≥ 20 yr (14,23,56,85,117,126,127) suggests that a high BMI and/or percent body fat may not be as important as predictors of heart disease as generally thought (51,81,91,92). One plausible explanation is the impact of regional fat distribution on cardiovascular risk factors (71). The deleterious effects of abdominal visceral fat are well established (71,107). On the other hand, thigh fat appears to be associated with a serum lipid profile that may reduce risk of atherosclerosis (107,133). In a study of 263 men and women, Terry et al. (133) found that thigh fat correlated positively with plasma high-density lipoprotein cholesterol (high density lipoprotein (HDL)-C) in women and negatively with plasma triglycerides in men. They concluded that "thigh fat may contribute to lipoprotein profiles that predict lower risk of cardiovascular disease." Pouliot et al. (107) reported similar findings on 58 overweight men. In these two studies the correlations between thigh fat and either plasma HDL-C or triglycerides were of the same magnitude (∼0.30-0.50)-but in the opposite direction-as those demonstrated for visceral abdominal fat and either plasma HDL-C or triglycerides. That a particular body fat depot could be considered beneficial to cardiovascular health obviously runs counter to conventional wisdom. However, it does offer a measure of support for the prophetic statement by Andres (3), who, after reviewing the literature on obesity and total mortality nearly 20 years ago, concluded that "not only does advice on the subject of obesity need reappraisal but that research into the possible associated benefits of moderate obesity would be worthwhile."
The apparently counterbalancing effects of thigh fat and abdominal visceral fat may help to explain the relatively low correlations between total body fat and blood lipids (71,144), as well as the fact that most autopsy and angiography studies of the putative risk factors for heart disease fail to show obesity or BMI to be a significant predictor of atherosclerosis (7,34,38,62,83,100). Interestingly, one of the largest of the case-control angiography studies (7) indicated that risk for coronary vessel disease decreased as BMI increased (i.e., a high BMI was protective).
EXCESS MORTALITY AMONG THIN INDIVIDUALS
A number of studies have reported increased mortality rates among thin men and women (27,29,31,39,42,43,48,53,78,131,137,138,140,142,147,148). The high death rates reported for individuals in the low BMI strata have been attributed to methodological flaws such as failure to control for smoking and the possibility that thinness may have been the consequence of weight loss caused by subclinical disease (80). These methodological flaws, however, have been ruled out as possible confounders in several epidemiological studies that have examined the BMI-mortality relationship separately for never smokers and/or nonsmokers (i.e., never smokers + ex-smokers), and have excluded from the statistical analysis deaths that occurred within the first several years of follow-up (27,29,31,39,113,114,131,137,140). Even with these added methodological safeguards, relatively high mortality rates have been observed among thin men and women.
In a 12-yr follow-up of 17,159 Finnish women ages 25-79 yr, Rissanen et al. (114) found that nonsmokers in the lowest BMI strata (BMI <20.6-24.0, depending upon age) consistently had higher mortality rates from any and all causes (relative risks ∼1.5-2.3, P < 0.05) compared with nonsmoking women in the referent group (BMI range = 20.7-26.4, depending upon age). Antecedent disease could not account for the results because the high mortality rate among thin women held true even after exclusion of deaths during the first 7 yr of follow-up. The authors noted that "the ill-effects of thinness were evident not only among women who were extremely thin but also among those whose BMI was well within the range customarily considered as acceptable (20-25 kg·m−2). Similar results were reported for a 13-yr follow-up of 20,015 Finnish men (113).
A meta-analysis of the BMI mortality literature by Troiano et al. (137) revealed increased mortality at moderately low BMI (<23) for nonsmoking white middle-aged men followed for 30 yr. The mortality rate associated with a BMI of 20 was essentially the same as that associated with a BMI of 30, suggesting that a body weight within the recommended range (2,89) was as detrimental to longevity as extreme overweight. Troiano et al. (137) cautioned that "attention to the health risks of underweight is needed, and body weight recommendations for optimum longevity need to be considered in light of these risks."
Why thinness would increase mortality rates is unknown. The high mortality rates observed in thin men and women cannot be attributed to smoking or to disease present at entry into the study (27,29,31,39,113,114,131,137,140). Nor do they necessarily reflect extreme thinness because the high death rates among thin men and women have been observed within the BMI range considered healthy. Increased risk of death caused by complications resulting from osteoporosis-linked fractures may account for some of the excess mortality in thin women (9). Low serum levels of vitamin A and carotene have been suggested as a possible explanation for increased cancer risk in thin men (39). Thin men and women may lack the nutritional reserves of obese men and women, making it less likely that they survive acute illness (106). Total mortality was inversely related to fatness level among men in the West of Scotland Prospective Cohort Study (39). Among never-smoking women ages 65-74 in the Epidemiological Follow-up study of the first National Health and Nutrition Examination Survey (26), thin, lean women had a death rate 3.1 times that of relatively fat (≥85th percentile of skinfolds) women within the mid-range of BMI (15-84th percentile).
Behavioral factors may help to explain some the excess mortality associated with thinness. Data from the 1990 Behavioral Risk Factor Surveillance System (120), for example, revealed that thin men (BMI < 20.7) and women (BMI < 19.1) were more likely to be sedentary and consume more dietary fat than men and women classified as normal weight (BMI = 20.8-27.8 for men; BMI = 19.2-27.3 for women). Thus for some persons low BMI may reflect an unhealthy lifestyle, which may contribute to a higher mortality rate. Physical inactivity, and the low aerobic fitness level that can be expected to accompany a sedentary lifestyle, increase mortality rates independently of BMI (11,17,95). In the Aerobics Center Longitudinal Study, thin men and women with low aerobic fitness levels had the highest mortality rates (17).
WEIGHT LOSS AND MORTALITY
Weight loss is generally recommended for the overweight primarily because weight loss intervention frequently is associated with improvements in health status (155). Theoretically, weight loss in overweight individuals should reduce death rates. However, there is scant evidence available to support this scenario (152,153). In fact, most relevant epidemioligical studies indicate that weight loss is associated with increased mortality rates (5,18,34,38,42,43,47,70,77,78,95,97,98,148).
In the NHANES I Epidemiologic Follow-up Study, Pamuk et al. (98) reported that weight loss of > 5% of maximum lifetime body weight was associated with increased death rates, particularly from cardiovascular disease. This was true for men with BMI <29, and for women in all BMI strata. Most notably, among women with maximum lifetime BMI between 26 and 29, weight loss >5% of maximum body weight was associated with relative risks for cardiovascular disease mortality that were 2.3-3.6 times that of the referent group (maximum weight loss < %5 of maximum body weight). The higher death rates associated with weight loss were observed in never smokers and after excluding from the data analysis participants who died within the first 5-8 yr of follow-up (thus minimizing potential confounding from antecedent disease). Pamuk et al. (97) opined that, "The policy of recommending a return to 'ideal' weight-by definition, a BMI below 26-to persons who are already overweight may need to be reconsidered." An estimated 40.3% of U.S. men and 58.5% of U.S. women with BMI 26-29.9 may be trying to lose weight (154).
Weight loss also was predictive of higher mortality rates in the Framingham Study (48), the Harvard Alumni Health Study (70), the Multiple Risk Factor Intervention Trial (MRFIT) (18), and a number of other epidemiological studies (for review, see (5)). Among Framingham men, for example, weight loss was associated with significantly higher (RR = 1.33-1.61, compared with weight stable group) 20-yr mortality from all causes, cardiovascular disease, and CHD. Weight gain, on the other hand, was not associated with increased mortality rates (RR = 0.78-0.86), despite the fact that during the study increases in blood pressure were greatest in the group of men who gained weight (48). The finding that moderate weight gain during adult years was not predictive of increased mortality has been reported by others as well (5,81,153).
The higher mortality rates associated with weight loss, especially from cardiovascular disease, among men in MR-FIT is surprising in view of the fact that these men were at high risk for CHD before entering the trial. Yet in every statistical comparison involving weight loss reported in this study (32 comparisons total), weight loss was associated with increased risk of death during the follow-up (range of RR = 1.04-3.42). Moderately overweight men at high risk for heart disease might be expected to be the optimum beneficiaries of weight loss. Yet among MRFIT men with baseline BMI between 26.08 and 28.82, those who lost >5% of body weight had a cardiovascular disease mortality rate that was 195% higher than the referent group of men (< 5% weight change from baseline).
A major limitation of nearly all epidemiological studies on weight loss and mortality is that they do not distinguish between intentional and unintentional weight loss (152). The American Cancer Society's Cancer Prevention Study I, however, addressed the issue of intentionality (153). Among 43,457 never-smoking overweight (BMI > 27) U.S. white women, ages 40-64, the impact of intentional weight loss on mortality rates depended upon health status. Among the 15,069 women who had one or more obesity-related health conditions, intentional weight loss was associated with a ∼20% reduction in all-cause mortality over a 12-yr follow-up. It is noteworthy that the reduction in death rate was independent of the amount of weight lost: women who lost 1-19 1b experienced the same reduction in all-cause mortality rate as did women who lost 20 or more pounds. This suggests that specific weight loss goals for overweight women with obesity-related health conditions may not be necessary.
Among the 28,388 women with no pre-existing illness, intentional weight loss generally was not associated with reduced mortality rates. To the contrary, in some instances intentional weight loss was associated with statistically significantly higher mortality rates. For example, intentional weight loss of 1-19 1b that occurred over a period of ≥ 1 yr was associated with increased all-cause (RR = 1.40) and cardiovascular disease (RR = 1.70) mortality. Thus, blanket recommendations for all overweight persons to lose weight, regardless of current health status, may not be justified.
A more recent report from the Cardiovascular Health Study (29) indicated that among older (ages 65-100), over-weight (mean BMI = 26.2-27.3) men and women, those who intentionally lost ≥ 10 1b during the year before entering the study had 5-yr mortality rates that were not materially different from those who remained weight stable.
WEIGHT LOSS PARADOX
The high mortality rates associated with weight loss, particularly for cardiovascular disease, are paradoxical in view of the conventional wisdom on weight loss. Weight loss usually is associated with improvements in established cardiovascular risk factors, and this theoretically should translate into reduced mortality. Not only does the bulk of epidemiological evidence fail to support the "lose weight-live longer" paradigm, it suggests that weight loss in some instances may do more harm than good, particularly for persons with no pre-existing health conditions. A number of possibilities may help to explain the weight loss paradox.
Intentional weight loss is not always associated with an improved health profile. Low carbohydrate diets, for example, increase serum total cholesterol (111) and low-density lipoprotein cholesterol (low density lipoprotein (LDL)-C) (68) and reduce serum HDL-C, particularly in women (68). These unfavorable cholesterol changes occurred despite reported mean weight losses of 3.1-7.7 kg (68,111). Even if the diets are not maintained, the transient deterioration in serum lipoproteins may influence the atherosclerotic process (25). Furthermore, low-carbohydrate diets typically involve increased consumption of animal products (68,111). Dietary intake of animal products is associated with higher levels of arachidonate in circulating blood lipids (101,102), and the absolute level of arachidonate in red cell membranes has been reported to be positively correlated to CHD risk (146). This may have particular relevance to the U.S. population because low-carbohydrate diets have been the most popular weight-reducing diets in the United States since the early 1960s.
Major weight loss has been reported to reduce adipose tissue and serum concentration of α-linolenic acid (103,130), a reduction that apparently cannot be prevented even with dietary α-linolenic acid supplementation (130). α-linolenic acid, and longer-chain w3 fatty acids for which it serves as a metabolic precursor (i.e., eicosapentaenoic acid; docosahexaenoic acid), are associated with reduced risk of atherosclerosis (64). Tang et al. (130) suggested that "a subtle but chronic risk state could be established if recurrent dieting depletes w3 reserves and intake during maintenance or weight gain does not allow effective repletion."
Adverse health outcomes associated with weight loss may also be attributable in part to weight loss drugs (26,82,92), including increased risk for primary pulmonary hypertension (82) and valvular heart disease (26). Because the use of weight loss drugs is disproportionately higher among over-weight and obese individuals, the increased cardiovascular disease morbidity and mortality associated with high BMI may reflect in part the effects of pharmacotherapy and not necessarily obesity per se (34).
WEIGHT CYCLING AND HEALTH RISK
It is possible that the increased mortality rates associated with weight loss are caused by weight cycling and not necessarily weight loss itself. Several epidemiological studies revealed higher all-cause (primarily from cardiovascular disease) death rates among men and women who had experienced considerable weight fluctuation (18,41,70,76,77).
The higher cardiovascular disease death rates associated with weight fluctuation may reflect elevated blood pressure that has been documented to occur during repeated cycles of weight loss and regain (34,35,122,123,149). Observations during World War II (21,58) revealed that hypertensive episodes could occur during refeeding subsequent to prolonged and severe caloric restriction. More recent studies on humans also provide evidence of increased cardiovascular risk associated with weight regain subsequent to weight loss (46,104).
The increased morbidity and mortality associated with weight cycling in humans is supported by a number of animal models (24,30,34,35,45,73,108,122,123,129,139,149). Weight cycling has been reported to induce hypertension and vascular damage in mice (123), rats (34,35), swine (122), and dogs (149). Increased sympathetic activity during refeeding has been proposed as the cause of the weight cycling-induced hypertensive responses (35). In rats, weight cycling has been shown to increase plasma insulin levels (73). Elevated plasma insulin is a significant predictor of cardiovascular disease mortality (37) and is a hallmark of the insulin resistance syndrome (71). Weight cycling in humans has been reported to be associated with increased risk of type 2 diabetes (49).
Weight cycling has also been reported to significantly modify whole body fatty acid composition in rats (i.e., decreases in linoleate and α-linolenate, and increases in myristate, palmitate, and palmitoleate) (24). Reduced adipose tissue concentrations of linoleate and α-linolenate may increase risk of cardiovascular disease (112). Chen et al. (24) demonstrated that carcass and adipose tissue fatty acid compositional changes were similar whether weight cycling (two cycles) was induced by caloric intake reductions of 36, 60, or 100% of daily caloric intake, thus suggesting that a relatively moderate caloric restriction is potentially as hazardous as complete fasting.
In addition to increasing cardiovascular risk, weight cycling may also increase risk for certain cancers (75,129,139). Although food restriction generally inhibits chemically induced tumorigenesis (63), Tagliaferro et al. (129) reported that cyclic food restriction increased mammary tumor incidence in rats. Additionally, Uhley at al. (139) reported that weight cycling increased oxidative DNA damage levels in mammary glands of rats. Regardless of whether rats were weight cycled above or below baseline weight, 4-5 weight cycles of ± 20% of baseline weight increased levels of 5-hydroxymethyl-2′-deoxyuridine (5-OHmU) in mammary gland DNA. 5-OHmU represents one type of oxidative DNA damage and has been reported be a potentially useful marker for breast cancer in humans (30). Whether the high breast cancer rates in the United States are related to weight cycling is entirely speculative, but the animal data are certainly cause for concern given the prevalence of weight loss attempts by U. S. women (132). In a study of Swedish women, Lindblad et al. (75) found that weight cycling was a major predictor of renal cell cancer.
In contrast to the association between weight cycling and higher death rates reported in epidemiological studies, weight cycling generally is not associated with an elevation of cardiovascular disease risk factors (90,91,155). Wing et al. (155), in fact, suggested that weight cycling might be preferred to remaining overweight or obese, the rationale being that lifetime "cumulative risk" of cardiovascular disease would be less in weight cyclers because of the improved risk profile (e.g., reduced blood pressure and serum lipids) during periods of weight loss. However, the epidemiological evidence argues strongly against this hypothesis, as no study to date has shown weight cycling to be associated with reduced cardiovascular disease, or all-cause, mortality. Also, it cannot be assumed that all risk factors for cardiovascular disease improve with weight loss (68,101,103,111,130). Further, even though whole-body composition (i.e., % body fat) may not change with weight cycling (90,91,155), alterations in fatty acid composition of various tissues as a result of chronic weight fluctuation via dieting may have adverse consequences (24,101). Finally, it may not be appropriate to draw conclusions about the effect of weight cycling on blood pressure or serum lipids unless measurements are made during the active phase of regaining weight. Blood pressure changes, for example, are significantly correlated to the rate of change in body weight but not to the absolute change in body weight (36).
IMPROVED HEALTH AND REDUCED MORBIDITY/MORTALITY INDEPENDENT OF WEIGHT LOSS
Overweight/obesity-related comorbidities, particularly those associated with the insulin resistance syndrome (e.g., hypertension, dyslipidemias, and hyperinsulinemia) (71), can be improved rather independently of weight loss (6,8,12,13,15,19,28,33,44,65,67,72,109,134,145,157). The Dietary Approaches to Stop Hypertension trial (6), for example, demonstrated that blood pressures could be effectively lowered (to the extent typically observed with initiation of pharmacotherapy) with simple changes in diet in the absence of weight loss. Leserman et al. (72) demonstrated that reductions in blood pressure achieved by lifestyle intervention alone could be maintained for 3-5 yr despite significant increases in body weight. These studies suggest that behavioral factors are more important than relative body weight as determinants of blood pressure.
Reductions in blood pressure accompanying exercise training in obese persons appear to be more strongly linked to decreases in plasma insulin than to reductions in body fat (65). In this regard, reductions in body fat via exercise are not necessary for improving insulin sensitivity. Nearly 30 years ago Bjorntorp (15) demonstrated that exercise training in obese women and men could significantly improve insulin sensitivity even in the face of an increase in body fat. More recently, Lamarche et al. (67) have confirmed these findings. Even when weight loss does occur with combined exercise and diet intervention, it appears that exercise is more important than weight loss itself for improving insulin sensitivity. After 6-8 wk of combined exercise and diet therapy in obese men and women with type 2 diabetes, multiple regression analysis revealed that improvement in insulin sensitivity was significantly correlated with quantity of exercise (number of steps per day) but not with changes in body weight (157). That obesity per se is not the underlying cause of insulin resistance and hyperinsulinemia is also supported by the data of Barnard et al. (13), who reported that a 3-wk lifestyle modification program was able to normalize fasting plasma insulin levels and eliminate the need for medication in a number of obese men and women with insulin resistance and/or type 2 diabetes, despite the fact that the subjects lost relatively little weight and still could be classified as significantly overweight or obese at the end of the program.
Barnard (12) also demonstrated that reductions in serum lipids following exercise and diet modification are largely unrelated to decreases in body weight. Among 4587 men and women who underwent a 3-wk, residential lifestyle modification program emphasizing aerobic exercise and healthy eating, less than 3% of the reductions in serum total cholesterol (234 to 180 mg·dL−1), LDL-C (151 to 116 mg·dL−1) and triglycerides (200.2 to 135.2 mg·dL−1) could be explained by reductions in body weight. That quality of diet is more important than weight loss for improving serum lipid and lipoprotein profile is also supported by the data of Ehnholm et al. (33), who showed that diet-induced reductions in atherogenic serum lipids and lipoproteins observed in hypercholesterolemic men and women were not attributable to weight loss.
In overweight men, exercise training independent of weight loss has been reported to inhibit secondary platelet aggregability (109) and reduce fasting and postprandial triglyceride-rich lipoprotein levels (145). In the Cholesterol Lowering Atherosclerosis Study (19), modest reductions in dietary fat intake among moderately overweight men with heart disease prevented the formation of new atherosclerotic lesions during a 2-yr trial. This occurred in the absence of weight loss and is consistent with the results of other intervention trials (8,28). In the Leiden Intervention Trial (8), progression of existing coronary lesions in some patients was reduced by diet therapy that resulted in only minimal weight loss. The Lyon Diet Heart Study (28) showed that, among moderately overweight men and women who had survived a myocardial infarction, an α-linolenic acid-rich diet, in the absence of weight loss, reduced myocardial infarction and mortality rates during a 5-yr follow-up.
Weight loss does not appear to be necessary for the reduced mortality rates reported in men who increase either their aerobic fitness (16) or level of physical activity (96). Blair et al. (16) reported that, compared with unfit men who remained unfit, unfit men who improved their fitness level experienced a 44% reduction in all-cause mortality rate during a follow-up lasting an average of 5.1 yr. This reduction in mortality rate was independent of changes in body weight, and, it is worth noting, more than twice that (∼20%) reported to be associated with intentional weight loss (153). Data from this cohort also indicate that the apparent effect of aerobic fitness on reducing all-cause mortality rate increases linearly with BMI (11), thus suggesting that the most overweight/obese men stand to benefit the most from starting an exercise program to improve aerobic fitness. In a longitudinal study of Harvard Alumni (96), beginning moderately vigorous sports activity was associated with a 23% reduction in all-cause mortality rate. The reduction in death rate was independent of changes in BMI; in fact, weight loss in among men in this cohort is associated with an increased mortality rate (70).
The preponderance of epidemiological evidence does not support the hypothesis that lowest mortality rates are observed in thin men and women. To the contrary, a number of studies indicate that low BMI is associated with mortality rates greater than those observed in men and women with average or above-average BMI. All studies considered, epidemiological evidence published to date provides no clear-cut picture as to what constitutes a "best" body weight in terms of longevity. This is not to say that BMI is irrelevant with regard to health, as the extremes of the BMI distribution frequently are reported to be associated with excess mortality (38,74,121,137). But what constitutes "extreme," and precisely where to draw the line between "healthy/normal" and "underweight," "overweight" or "obese," is not wholly apparent from the literature. It may be more efficacious to adopt the suggestion of Knapp (60), who, after MetLife issued its most recent weight tables in 1983, urged that we focus less on so-called "ideal weights" and more on "dangerous weights," i.e., those at both extremes of the BMI distribution.
The majority of epidemiological evidence also fails to support the hypothesis that weight loss reduces mortality rates. However, it would be inappropriate to conclude that weight loss would not be potentially therapeutic for certain individuals, i.e., those with "dangerous weights" and the attendant obesity-related health problems more prevalent in this population. Truly overweight/obese individuals with comorbidities most likely will benefit from intentional weight loss (153). However, because the reduction in mortality rate appears to be independent of the amount of weight loss (153), specific weight loss goals may be unnecessary. Furthermore, increasing physical activity and aerobic fitness, independent of weight loss, has been shown to reduce mortality rates by a greater amount than intentional weight loss (23-44% vs 20%) (16,96,153). It is also quite evident that amelioration of many of the most common obesity-related comorbidities (hypertension, dyslipidemias, insulin resistance, glucose intolerance) can be achieved via lifestyle changes independently of weight loss. Thus, for many persons a higher-than-average BMI (or percent body fat) is perhaps more aptly viewed as a relatively imprecise marker for an unhealthy lifestyle, the metabolic manifestation of which might be the insulin resistance syndrome and its associated disorders (71). If insulin resistance and its metabolic sequelae are the true culprits in much of the morbidity and mortality attributed to overweight/obesity, public health may be better served by focusing more on increasing physical activity and fitness and on improving nutrition, than on weight reduction per se. Because of the rather poor success rates for permanent weight loss (132), this may be the only viable option for millions of over-weight/obese U. S. adults.
As for individuals who are overweight/obese by BMI criteria (2,66), but who have no weight-related health problems, weight loss certainly is not without risks (5,38,97,98). Thus, categorical recommendations for "overweight/obese" individuals to lose weight, regardless of current health status, may be irresponsible (115,156). For even if a "statistically desirable weight" (1) is identified in any of the various epidemiological studies, it does not necessarily follow that all individuals will be better off if they reduce (or gain, as the case may be) body mass to the "statistically best" weight. There may be people who are overweight or obese by BMI criteria who are at a natural healthy weight and who most likely are better off maintaining that weight. Urging naturally heavy men and women to lose weight, without also considering lifestyle factors and current health status, may be counterproductive and also may violate a major tenet of the helping professions, "above all else, do no harm." As Kassirer and Angell (55) cautioned, "until we have better data about the risks of being overweight and the benefits and risks of trying to lose weight, we should remember that the cure for obesity may be worse than the condition."
1. Abernathy, R. P. and D. R. Black. Healthy body weights: an alternative perspective. Am. J. Clin. Nutr.
2. American Dietetic Association. Position of the American Dietetic Association: Weight Management. J. Am. Diet. Assoc.
3. Andres, R. Effect of obesity on total mortality. Int. J. Obes.
4. Andres, R., D. Elahi, J. D. Tobin, D. C. Muller, and L. Brant. Impact of age on weight goals. Ann. Int. Med.
5. Andres, R., D. C. Muller, and J. D. Sorkin. Long-term effects of change in body weight on all-cause mortality: a review. Ann. Int. Med.
6. Appel, L. J., T. J. Moore, E. Obarzanek, et al. A clinical trial of the effects of dietary patterns on blood pressure. N. Engl. J. Med.
7. Applegate, W. B., J. P. Hughes, and R. V. Zwagg. Case-control study of CHD risk factors in the elderly. J. Clin. Epidemiol.
8. Arntzenius, A. C., D. Kromhout, J. D. Barth, et al. Diet, lipoproteins, and the progression of coronary atherosclerosis: the Leiden Intervention Trial. N. Engl. J. Med.
9. Avioli, L. V. Significance of osteoporosis: a growing international health care problem. Calcif. Tissue Int.
10. Ballard-Barbash, R. and C. A. Swanson. Body weight: estimation of risk for breast and endometrial cancers. Am. J. Clin. Nutr.
11. Barlow, C. E., H. W. Kohl, III, L. W. Gibbons, and S. N. Blair. Physical fitness, mortality and obesity. Int. J. Obes.
19(Suppl. 4):S41-S44, 1995.
12. Barnard, R. J. Effects of life-style modification on serum lipids. Arch. Int. Med.
13. Barnard, R. J., E. J. Ugianskis, D. A. Martin, and S. B. Inkeles. Role of diet and exercise in the management of hyperinsulinemia and associated atherosclerotic risk factors. Am. J. Cardiol.
14. Barrett-Connor, E. L. Obesity, atherosclerosis, and coronary artery disease. Ann. Int. Med.
15. Bjorntorp, P., K. de Jounge, L. Sjostrom, and L. Sullivan. The effect of physical training on insulin production in obesity. Metabolism
16. Blair, S. N., H. W. Kohl III, C. E. Barlow, R. S. Paffenbarger, L. W. Gibbons, and C. A. Macera. Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men. JAMA
17. Blair, S. N., H. W. Kohl III, R. S. Paffenbarger, D. G. Clark, K. H. Cooper, and L. W. Gibbons. Physical fitness and all-cause mortality: a prospective study of healthy men and women. JAMA
18. Blair, S. N., J. Shaten, K. Brownell, G. Collins, and L. Lissner. Body weight change, all-cause mortality, and cause-specific mortality in the Multiple Risk Factor Intervention Trial. Ann. Int. Med.
19. Blankenhorn, D. H., R. L. Johnson, W. J. Mack, H. A. El Zein, and L. I. Vailas. The influence of diet on the appearance of new lesions in human coronary arteries. JAMA
20. Borhani, N. O., H. H. Hechter, and L. Breslow. Report of a ten-year follow-up study of the San Francisco longshoreman. J. Chronic Dis.
21. Brozek, J., C. B. Chapman, and A. Keys. Drastic food restriction: effect on cardiovascular dynamics in normotensive and hypertensive conditions. JAMA
22. Society of Actuaries and Association of Life Insurance Medical Directors of America. Build Study 1979.
Chicago: Soc. of Actuaries and Assoc. of Life Insurance Medical Directors of America, 1980, pp. 1-255.
23. Carmelli, D., J. Halpern, G. E. Swan, et al. 27-year mortality in the Western Collaborative Group Study: construction of risk groups by recursive partitioning. J. Clin. Epidemiol.
24. Chen, Z.-Y., M.-M. Sea, K.-Y. Kwan, Y.-H. Leung, and P.-F. Leung. Depletion of linoleate induced by weight cycling is in-dependent of extent of calorie restriction. Am. J. Physiol.
272 (Regulatory Integr. Comp:Physiol. 41):R43-R50, 1997.
25. Cleland, J. G. F. and D. M. Krikler. Modification of atherosclerosis by agents that do not lower cholesterol. Br. Heart J.
26. Connolly, H. M., J. L. Crary, M. D. McGoon, et al. Valvular heart disease associated with fenfluramine-phentermine. N. Engl. J. Med.
27. Cornoni-Huntley, J. C., T. B. Harris, D. F. Everett, et al. An overview of body weight of older persons, including the impact on mortality: The National Health and Nutrition Examination Survey I-epidemiologic follow-up study. J. Clin. Epidemiol.
28. de Lorgeril, M., S. Renaud, N. Mamelle, et al. Mediterranean alpha-linolenic acid-rich diet in secondary prevention of CHD. Lancet
29. Diehr, P., D. E. Bild, T. B. Harris, A. Duxbury, D. Siscovick, and M. Rossi. Body mass index and mortality in nonsmoking older adults: The Cardiovascular Health Study. Am. J. Public Health
30. Djuric, Z., L. K. Heilbrun, M. S. Simon, et al. Levels of 5-hydroxymethyl-2′-deoxyuridine in DNA from blood as a marker of breast cancer. Cancer
31. Dyer, A. R., J. Stamler, D. M. Berkson, and H. A. Lindberg. Relationship of relative weight and body mass index to 14-year mortality in the Chicago Peoples Gas Company Study. J. Chronic Dis.
32. Edelstein, S. L. and E. Barrett-Connor. Relation between body size and bone mineral density (BMD) in elderly men and women. Am. J. Epidemiol.
33. Ehnholm, C., J. K. Huttunen, P. Pietinen, et al. Effect of diet on serum lipoproteins in a population with a high risk of CHD. N. Engl. J. Med.
34. Ernsberger, P. and P. Haskew. Health implications of obesity: an alternative view. J. Obes. Weight Regul.
35. Ernsberger, P., R. J. Koletsky, J. S. Baskin, and L. A. Collins. Consequences of weight cycling in obese spontaneously hypertensive rats. Am. J. Physiol.
270 (Regulatory Integr. Comp: Physiol.39): R864-R872, 1996.
36. Ernsberger, P. and D. O. Nelson. Effects of fasting and refeeding on blood pressure are determined by nutritional state, not by body weight change. Am. J. Hypertens.
37. Fontebonne, A., M. A. Charles, N. Thibult, et al. Hyperinsulinemia as a predictor of CHD mortality in a healthy population: The Paris Prospective Study, 15-year follow-up. Diabetologia
38. Gaesser, G. A. Big Fat Lies: The Truth About Your Weight and Your Health.
New York: Fawcett Columbine, 1996, pp. 1-312.
39. Garn, S. M., V. M. Hawthorne, J. J. Pilkington, and S. D. Pesick. Fatness and mortality in the West of Scotland. Am. J. Clin. Nutr.
40. Gordon, T. and J. T. Doyle. Weight and mortality in men: The Albany Study. Int. J. Epidemiol.
41. Hamm, P., R. B. Shekelle, and J. Stamler. Large fluctuations in body weight during young adulthood and twenty-five-year risk of coronary death in men. Am. J. Epidemiol.
42. Harris, T. B., R. Ballard-Barbasch, J. Madans, D. M. Makuc, and J. J. Feldman. Overweight, weight loss, and risk of CHD in older women: The NHANES I Epidemiologic Follow-up Study. Am. J. Epidemiol.
43. Harris, T., E. F. Cook, R. Garrison, M. Higgins, W. Kannel, and L. Goldman. Body mass index and mortality among nonsmoking older persons: The Framingham Heart Study. JAMA
44. Hellenius, M.-L., I. Krakau, and U. de Faire. Favourable long-term effects from advice on diet and exercise given to healthy men with raised cardiovascular risk factors. Nutr. Metab. Cardiovasc. Dis.
45. Hembrough, F. B. and D. H. Riedesel. Mechanical behavior change in a major artery after a series of starvation-refeeding episodes. Am. J. Physiol.
46. Hewing, R., H. Leibermeister, H. Daweke, F. A. Gries, and D. Gruneklee. Weight regain after low calorie diet: long-term pattern of blood sugar, serum lipids, ketone bodies, and serum insulin levels. Diabetologia
47. Higgins, M., R. D'Agostino, W. Kannel, and J. Cobb. Benefits and adverse effects of weight loss: observations from the Framingham Study. Ann. Int. Med.
48. Hoes, A. W., D. E. Grobbee, H. A. Valkenburg, J. Lubsen, and A. Hofman. Cardiovascular risk and all-cause mortality: a 12 year follow-up study in The Netherlands. Eur. J. Epidemiol.
49. Holbrook, T. L., E. Barrett-Connor, and D. L. Wingard. The association of lifetime weight and weight control patterns with diabetes among men and women in an adult community. Int. J. Obes.
50. Hopkins, P. N. and R. R. Williams. A survey of 246 suggested coronary risk factors. Atherosclerosis
51. Hubert, H. B., M. Feinleib, P. M. McNamara, and W. P. Castelli. Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation
52. Hunter, D. J. and W. C. Willett. Diet, body size, and breast cancer. Epidemiol. Rev.
53. Jarrett, R. J., M. J. Shipley, and G. Rose. Weight and mortality in the Whitehall Study. Br. Med. J.
54. Kabat, G. C. and E. L. Wynder. Body mass index and lung cancer risk. Am. J. Epidemiol.
55. Kassirer, J. P. and M. Angell. Losing weight: an ill-fated New Year's resolution. N. Engl. J. Med.
56. Keil, J. E., S. E. Sutherland, R. G. Knapp, D. T. Lackland, P. C. Gazes, and H. A. Tyroler. Mortality rates and risk factors for coronary disease in black as compared with white men and women. N. Engl. J. Med.
57. Keys, A. Overweight, obesity, CHD, and mortality. Nutr. Rev.
58. Keys, A., J. Brozek, A. Henschel, O. Michelson, and H. Longhurst. Biology of Human Starvation.
Minneapolis, MN: Univ. Minn. Press, 1950, pp. 618-622.
59. Keys, A., C. Aravanis, H. Blackburn, et al. Serum cholesterol and cancer mortality in the Seven Countries Study. Am. J. Epidemiol.
60. Knapp, T. R. A methodological critique of the "ideal weight" concept. JAMA
61. Knekt, P., M. Heliovaara, A. Rissanen, et al. Leanness and lung-cancer risk. Int. J. Cancer
62. Kramer, J. R., Y. Matsuda, J. C. Mulligan, M. Aronow, and W. L. Proudfit. Progression of coronary atherosclerosis. Circulation
63. Kritchevsky, D., C. B. Welch, and D. M. Klurfield. Response of mammary tumors to caloric restriction for different time periods during the promotion phase. Nutr. Cancer
64. Kromhout, D., E. B. Bosschieter, and C. de Coulander. The inverse relation between fish consumption and 20-year mortality from CHD. N. Engl. J. Med.
65. Krotkiewski, M., K. Mandroukas, L. Sjostrom, L. Sullivan, H. Wetterqvist, and P. Bjorntorp. Effects of long-term physical training on body fat, metabolism, and blood pressure in obesity. Metabolism
66. Kuczmarski, R. J., K. M. Flegal, S. M. Campbell, and C. L. Johnson. Increasing prevalence of overweight among U.S. adults: The National Health and Nutrition Examination Surveys, 1960-1991. JAMA
67. Lamarche, B., J.-P. Despres, M.-C. Pouliot, et al. Is body fat loss a determinant factor in the improvement of carbohydrate and lipid metabolism following aerobic exercise training in obese women? Metabolism
68. Larosa, J. C., A. G. Fry, R. Muesing, and D. R. Rosing. Effects of high-protein, low-carbohydrate dieting on plasma lipoproteins and body weight. J. Am. Diet. Assoc.
69. Lee, I.-M., J. E. Manson, C. H. Hennekens, and R. S. Paffenbarger. Body weight and mortality: a 27-year follow-up of middle-aged men. JAMA
70. Lee, I.-M. and R. S. Paffenbarger. Change in body weight and longevity. A 27-year follow-up of middle-aged men. JAMA
71. Lemieux, S. and J.-P. Despres. Metabolic complications of visceral obesity: contribution to the aetiology of type 2 diabetes and implications for prevention and treatment. Diabete Metab. (Paris)
72. Leserman, J., E. M. Stuart, M. E. Mamish, et al. Nonpharmacologic intervention for hypertension: long-term follow-up. J. Cardiopulm. Rehabil.
73. Levin, B. Diet cycling and age alter weight gain and insulin levels in rats. Am. J. Physiol.
267(Regulatory Integr. Comp Physiol.36):R527-R535, 1994.
74. Lew, E. A. and L. Garfinkel. Variations in mortality by weight among 750,000 men and women. J. Chronic Dis.
75. Lindblad, P., A. Wolk, R. Bergstrom, I. Persson, and H.-O. Adami. The role of obesity and weight fluctuations in the etiology of renal cell cancer: a population-based case-control study. Cancer Epidemiol. Biomarkers Prev.
76. Lissner, L., C. Bengtsson, L. Lapidus, B. Larsson, B. Bengtsson, and K. Brownell. Body weight variability and mortality in the Gothenburg prospective studies of men and women. In: Obesity in Europe 88,
P. Bjorntorp and S. Rössner (Eds.). London: John Libbey, 1989, pp. 55-60.
77. Lissner, L., P. M. Odell, R. B. D'Agostino, et al. Variability of body weight and health outcomes in the Framingham population. N. Engl. J. Med.
78. Losonczy, K. G., T. B. Harris, J. Cornoni-Huntley, et al. Does weight loss from middle age to old age explain the inverse weight mortality relation in old age? Am. J. Epidemiol.
79. Manson, J. E. and G. A. Faich. Pharmacotherapy for obesity: do the benefits outweigh the risks? N. Engl. J. Med.
80. Manson, J. E., M. J. Stampfer, C. H. Hennekens, and W. C. Willett. Body weight and longevity: a reassessment. JAMA
81. Manson, J. E., W. C. Willett, M. J. Stampfer, et al. Body weight and mortality among women. N. Engl. J. Med.
82. McCann, U. D., L. S. Seiden, L. J. Rubin, and G. A. Ricaurte. Brain serotonin neutrotoxicity and primary pulmonary hypertension from fenfluramine and dexfenfluramine: a systematic review of the evidence. JAMA
83. McGill, H. C., J. Arias-Stella, L. M. Carbonell, et al. General findings of the International Atherosclerosis Project. Lab. Invest.
84. McGinnes, J. M. and W. H. Foege. Actual causes of death in the United States. JAMA
85. Menotti, A., A. Keys, D. Kromhout, et al. Inter-cohort differences in CHD mortality in the 25-year follow-up of the Seven Countries Study. Eur. J. Epidemiol.
86. Metropolitan Life Insurance Company. Ideal weights for women. Stat. Bull.
87. Metropolitan Life Insurance Company. Ideal weights for men. Stat. Bull.
88. Metropolitan Life Insurance Company. New weight standards for men and women. Stat. Bull.
89. Metropolitan Life Insurance Company. Metropolitan height and weight tables. Stat. Bull.
90. Muls, E., K. Kempen, G. Vansant, and W. Saris. Is weight cycling detrimental to health? A review of the literature in humans. Int. J. Obes.
91. National Task Force on the Prevention and Treatment of Obesity. Weight Cycling. JAMA
92. National Task Force on the Prevention and Treatment of Obesity. Long-term pharmacotherapy in the management of obesity. JAMA
93. Norgan, N. G. The beneficial effects of body fat and adipose tissue in humans. Int. J. Obes.
94. Paffenbarger, R. S. Contributions of epidemiology to exercise science and cardiovascular health. Med. Sci. Sports Exerc.
95. Paffenbarger, R. S., R. T. Hyde, A. L. Wing, and C.-C. Hsieh. Physical activity, all-cause mortality, and longevity of college alumni. N. Engl. J. Med.
96. Paffenbarger, R. S., R. T. Hyde, A. L. Wing, I.-M. Lee, D. L. Jung, and J. B. Kampert. The association of changes in physical-activity level and other lifestyle characteristics with mortality among men. N. Engl. J. Med.
97. Pamuk, E. R., D. F. Williamson, J. Madans, M. K. Serdula, J. C. Kleinman, and T. Byers. Weight loss and mortality in a national cohort of adults, 1971-1987. Am. J. Epidemiol.
98. Pamuk, E. R., D. F. Williamson, M. K. Serdula, J. Madans, and T. E. Byers. Weight loss and subsequent death in a cohort of U.S. adults. Ann. Int. Med.
99. Pate, R. R., M. Pratt, S. N. Blair, et al. Physical activity and public health: a recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA
100. Patel, Y. C., D. A. Eggen, and J. P. Strong. Obesity, smoking and atherosclerosis: a study of interassociations. Atherosclerosis
101. Phinney, S. D. Weight cycling and cardiovascular risk in obese men and women. Am. J. Clin. Nutr.
102. Phinney, S. D., R. S. Odin, S. B. Johnson, and R. T. Holman. Reduced arachidonate in serum phospholipids and cholesteryl esters associated with vegetarian diets in humans. Am. J. Clin. Nutr.
103. Phinney, S. D., A. B. Tang, S. B. Johnson, and R. T. Holman. Reduced adipose tissue 18:3w3 with weight loss by very low calorie dieting. Lipids
104. Pfohl, M., D. Luft, I. Blomberg, and R.-M. Schmulling. Long-term changes in body weight and cardiovascular risk factors after weight reduction with group therapy and dexfenfluramine. Int. J. Obes.
105. Polivy, J. Psychological consequences of food restriction. J. Am. Diet. Assoc.
106. Potter, J., K. Klipstein, J. J. Reilly, and M. Roberts. The nutritional status and clinical course of acute admissions to a geriatric unit. Age Aging
107. Pouliot, M.-C., J.-P. Despres, A. Nadeau, et al. Visceral obesity in men: associations with glucose tolerance, plasma insulin, and lipoprotein levels. Diabetes
108. U. S. Public Health Service. Obesity and Health. Publ. No. 1485.
Washington D.C.: U.S. Government Printing Office, 1966, p.40.
109. Rauramaa, R., J. T. Salonen, K. Seppanen, et al. Inhibition of platelet aggregability by moderate-intensity physical exercise: a randomized clinical trial in overweight men. Circulation
110. Ribot, C., F. Tremollieres, J.-M. Pouilles, M. Bonneu, F. German, and J.-P. Louvet. Obesity and postmenopausal bone loss: the influence of obesity on vertebral density and bone turnover in postmenopausal women. Bone
111. Rickman, F., N. Mitchell, J. Dingman, and J. E. Dalen. Changes in serum cholesterol during the Stillman Diet. JAMA
112. Riemersma, R. A., D. A. Wood, S. Butler, et al. Linoleic acid content in adipose tissue and CHD. Br. Med. J.
113. Rissanen, A., M. Heliovaara, P. Knekt, A. Aromaa, and A. Reunanen. Overweight and mortality in Finnish men. In: Obesity in Europe 88,
P. Bjorntorp and S Rossner Eds.). Paris: John Libbey, 1989, pp. 61-68.
114. Rissanen, A., P. Knekt, M. Heliovaara, A. Aromaa, A Reunanen, and J Maatela. Weight and mortality in Finnish women. J. Clin. Epidemiol.
115. Robison, J. I., S. L. Hoerr, J. Strandmark, and B. Mavis. Obesity, weight loss, and health. J. Am. Diet. Assoc.
116. Seeman, E., L. J. Melton, III, W. M. O'Fallon, and B. L. Riggs. Risk factors for spinal osteoporosis in men. Am. J. Med.
117. Seidell, J. C., R. Andres, J. D. Sorkin, and D. C. Muller. The sagittal waist diameter and mortality in men: the Baltimore Longitudinal Study on Aging. Int. J. Obes.
118. Serdula, M. K., M. E. Collins, D. F. Williamson, R. F. Anda, E. Pamuk, and T. E. Byers. Weight control practices of U.S. adolescents and adults. Ann. Int. Med.
119. Serdula, M. K., D. F. Williamson, R. F. Anda, A. Levy, A. Heaton, and T. Byers. Weight control practices in adults: results of a multistate telephone survey. Am. J. Public Health
120. Simoes, E. J., T. Byers, R. J. Coates, M. K. Serdula, A. H. Mokdad, and G. W. Heath. The association between leisure-time physical activity and dietary fat in American adults. Am. J. Public Health
121. Sjostrom, L. V. Mortality of severely obese subjects. Am. J. Clin. Nutr.
122. Smith, G. S., J. L. Smith, M. S. Mameesh, J. Simon, and B. C. Johnson. Hypertension and cardiovascular abnormalities in starved-refed swine. J. Nutr.
123. Smith-Vaniz, G. T., A. D. Ashburn, and W. L. Williams. Diet-induced hypertension and cardiovascular lesions in mice. Yale J. Biol. Med.
124. Stehbens, W. E. Limitations of the epidemiological method in CHD. Int. J. Epidemiol.
125. Stevens, J., J. Cai, E. R. Pamuk, D. F. Williamson, M. J. Thun, and J. L. Woods. The effect of age on the association between body-mass index and mortality. N. Engl. J. Med.
126. Stevens, J., J. E. Keil, P. F. Rust, H. A. Tyroler, C. E. Davis, and P. C. Gazes. Body mass index and body girths as predictors of mortality in black and white women. Arch. Int. Med.
127. Stevens, J., J. E. Keil, P. F. Rust, et al. Body mass index and body girths as predictors of mortality in black and white men. Am. J. Epidemiol.
128. Suadicani, P., H. O. Hein, and F. Gyntelberg. Height, weight, and risk of colorectal cancer: an 18-year follow-up in a cohort of 5249 men. Scand. J. Gastroenterol.
129. Tagliaferro, A. R., A. M. Ronan, L. D. Meeker, H. J. Thompson, A. L. Scott, and D. Sinha. Cyclic food restriction alters substrate utilization and abolishes protection from mammary carcinogenesis in female rats. J. Nutr.
130. Tang, A. B., K. Y. Nishimura, and S. D. Phinney. Preferential reduction in adipose alpha-linolenic acid (18:3n-3) during very low caloric dieting despite supplementation with 18:3n-3. Lipids
131. Tayback, M., S. Kumanyika, and E. Chee. Body weight as a risk factor in the elderly. Arch. Intern. Med.
132. Technology Assessment Conference Panel. Methods for voluntary weight loss and control: Technology Assessment Conference Statement. Ann. Int. Med.
133. Terry, R. B., M. L. Stefanick, W. Haskell, and P. D. Wood. Contributions of regional adipose tissue depots to plasma lipoprotein concentrations in overweight men and women: possible protective effects of thigh fat. Metabolism
134. Tremblay A., J.-P. Despres, M.-C Pouliot, et al. Normalization of the metabolic profile in obese women by exercise and low fat diet. Med. Sci. Sports Exerc.
135. Tremollieres, F. A., J.-M. Pouilles, and C. Ribot. Vertebral postmenopausal bone loss is reduced in overweight women: a longitudinal study in 155 early postmenopausal women. J. Clin. Endocrinol. Metab.
136. Tretli, S. Height and weight in relation to breast cancer morbidity and mortality: a prospective study of 570,000 women in Norway. Int. J. Cancer
137. Troiano, R. P., E. A. Frongillo, Jr., J. Sobal, and D. A. Levitsky. The relationship between body weight and mortality: a quantitative analysis of combined information from existing studies. Int. J. Obes.
138. Tyroler, H. A., M. G. Knowles, S. B. Wing, et al. Ischemic heart disease risk factors and twenty-year mortality in middle-aged Evans County black males. Am. Heart J.
139. Uhley, V. E., M. A. Pellizzon, A. M. Buison, F. Guo, Z. Djuric, and K.-L. C. Jen. Chronic weight cycling increases oxidative DNA damage levels in mammary gland of female rats fed a high-fat diet. Nutr. Cancer
140. Vandenbroucke, J. P., B. J. Mauritz, A. de Bruin, J. H. H. Verheesen, C. van der Heide-Wessel, and R. M. van der Heide. Weight, smoking, and mortality. JAMA
141. Vatten, L. J. and S. Kvinnsland. Prospective study of height, body mass index and risk of breast cancer. Acta Oncol.
142. Waaler, H. T. Height, weight, and mortality: the Norwegian Experience. Acta Med. Scand. (Suppl.)
143. Wardlaw, G. M. Putting osteoporosis in perspective. J. Am. Diet. Assoc.
144. Weinsier, R. L., R. J. Fuchs, T. D. Kay, J. H. Triebwasser, and M. C. Lancaster. Body fat: its relationship to CHD, blood pressure, lipids and other risk factors measured in a large male population. Am. J. Med.
145. Weintraub, M. S., Y. Rosen, R. Otto, S. Eisenberg, and J. L. Breslow. Physical Exercise Conditioning in the absence of weight loss reduces fasting and postprandial triglyceride-rich lipoprotein levels. Circulation
146. Wenxun, F., R. Parker, B. Parpia, et al. Erythrocyte fatty acids, plasma lipids, and cardiovascular disease in rural China. Am. J. Clin. Nutr.
147. Wienpahl, J., D. R. Ragland, and S. Sidney. Body mass index and 15-year mortality in a cohort of black men and women. J. Clin. Epidemiol.
148. Wilcosky, T., J. Hyde, J. J. B. Anderson, S. Bangdiwala, and B. Duncan. Obesity and mortality in the Lipid Research Clinics Program: Follow-up Study. J. Clin. Epidemiol.
149. Wilhelmj, C. M., A. J. Carnazzo, and H. H. McCarthy. Effect of fasting and realimentation with diets high in carbohydrate or protein on blood pressure and heart rate of sympathectomized dogs. Am. J. Physiol.
150. Willett, W. C., J. E. Manson, M. J. Stampfer, et al. Weight, weight change, and CHD in women: risks within the "normal" weight range. JAMA
151. Williamson, D. F. Descriptive epidemiology of body weight and weight change in U.S. adults. Ann. Int. Med.
152. Williamson, D. F. and E. R. Pamuk. The association between weight loss and increased longevity: a review of the evidence. Ann. Int. Med.
153. Williamson, D. F., E. Pamuk, M. Thun, D. Flanders, T. Byers, and C. Heath. Prospective study of intentional weight loss and mortality in never-smoking overweight U.S. white women aged 40-64 years. Am. J. Epidemiol.
154. Williamson, D. F., M. K. Serdula, R. F. Anda, A. Levy, and T. Byers. Weight loss attempts in adults: goals, duration, and rate of weight loss. Am. J. Public Health
155. Wing, R. R., R. W. Jeffery, and W. L. Hellerstedt. A prospective study of effects of weight cycling on cardiovascular risk factors. Arch. Intern. Med.
156. Wooley, S. C. and D. M. Garner. Obesity treatment: the high cost of false hope. J. Am. Diet. Assoc.
157. Yamanouchi, K., T. Shinozaki, K. Chikada, et al. Daily walking combined with diet therapy is a useful means for obese NIDDM patients not only to reduce body weight but also to improve insulin sensitivity. Diabetes Care