RSS Feeds
Medical Hazards of Obesity
- From St. Luke's-Roosevelt Hospital Center and Columbia University College of Physicians and Surgeons, New York, New York. Requests for Reprints: F. Xavier Pi-Sunyer, MD, Obesity Research Center, St. Luke's-Roosevelt Hospital Center, New York, NY 10025. Grant Support: In part by grants P30-DK26687, DK40414, and R55-DK35911 from the National Institutes of Health.
Abstract
The medical hazards of obesity are discussed. Risks include insulin resistance, diabetes mellitus, hypertriglyceridemia, decreased levels of high-density lipoprotein cholesterol, and increased levels of low-density lipoprotein cholesterol. Obesity is also associated with gallbladder disease and some forms of cancer as well as sleep apnea, chronic hypoxia and hypercapnia, and degenerative joint disease. Obesity is an independent risk factor for death from coronary heart disease. A central distribution of body fat enhances the risk for most of these conditions.
To assess directly the health hazards of obesity, the magnitude of risk can only be measured through population studies [1]. The ways in which obesity affects health can be assessed by examining its effect on the duration of life and on the presence or onset of premature disease. Because body weight is determined by multiple factors including genetic, cultural, socioeconomic, behavioral, and situational mechanisms, it has been difficult to evaluate the individual contribution of each [2]. Covariables can have confounding effects that are extraordinarily difficult to dissect.
Several problems complicate the evaluation of the health hazards of obesity. First, the definition of obesity is imprecise. From an epidemiologic standpoint, dividing a population into an obese group and a nonobese group makes little sense without the demarkation of a threshold of weight at which the risk for death or disability occurs [1]. Second, current weight may be related to present health status or current health practices. Thus, a smoker tends to have a lower weight, but smoking affects his or her risks for morbidity and death [3]. Similarly, a person with a chronic disease (whether evident or not) may have a lower weight, but it is the disease rather than the weight that affects his or her risk for death. For this reason (among others), an underweight status is associated with higher risk [3].
Third, the duration of the follow-up can determine the prognostic importance of body weight [4]. Short-term studies (10 years or less) show little association between baseline relative weight and outcome. Most intermediate-range studies (10 to 20 years) show weak associations between body weight and outcome (although these associations vary with end point). Long-term studies ( 20 years), however, have reported a strong association between relative body weight and adverse outcome [46].
Fourth, the relation between obesity and disease outcome varies with the age of the study group. For example, analyses of cardiovascular end points among men indicate that obesity is a much more important predictor at younger than at older ages [58]. Obesity occurring early in life affects intervening risk factors such as diabetes and hypertension much more strongly than that occurring later in life, when these risk factors become more important variables [9].
Finally, most epidemiologic studies use a static approach. That is, people are placed in a weight group at the beginning of the study and are followed for the next 5 to 25 years classified in that weight category. Subsequent changes in weight and their relation to subsequent morbidity and mortality rates have only recently been addressed.
Specific Health Consequences
Obesity is associated with an increased risk for insulin resistance, hypertension, dyslipidemia, cardiovascular disease, non–insulin-dependent diabetes mellitus, gallstones and cholecystitis, respiratory dysfunction, and certain forms of cancer.
Blood Pressure
The association between hypertension and obesity is well documented. Cross-sectional studies have shown that obese persons have a greater risk for high blood pressure than do lean persons [7]. A cross-sectional study of the general population of the city of Bergen, Norway, showed that every 10-kg increase in body weight was associated with an increase of 3 and 2 mm Hg in systolic and diastolic pressures, respectively [10]. A similar increase was found in the Tecumseh study [11].
The Second National Health and Examination Survey (NHANES II), a cross-sectional study conducted from 1976 to 1980 on a representative sample of U.S. residents, showed the prevalence of hypertension among overweight adults to be 2.9 times that among nonoverweight adults [12]. The risk in persons 20 to 44 years old was 5.6 times greater than that in persons 45 to 74 years old, a figure which was two times higher than that for nonoverweight adults [13]. The data from NHANES II were calculated using as the reference population those persons who were at or below the 85th percentile of weight for their height. This cut-off point comes at a body mass index (BMI) of 27.8 in men and of 27.3 in women [12]. Body mass index is determined by dividing the weight in kilograms by the height squared in meters and is a good epidemiologic measure of obesity. Cross-sectional studies have shown that those persons who are 20% or more overweight have a prevalence of hypertension that is twice that among persons of normal weight [14]. In the Western world, about one third of cases of hypertension are thought to be due to obesity, and in men younger than 45 years this figure may reach 60% [15].
In the prospective, longitudinal study of residents of Framingham, Massachusetts, increases in weight were associated with increases in blood pressure. In men, for every 10% increase in relative weight, blood pressure increased by 6.5 mm Hg. A 15% gain in weight was associated with an 18% increase in systolic pressure [16]. Persons who were only 20% overweight had an eightfold greater incidence of hypertension [17]. Another prospective study in Georgia showed that a gain in weight during a 6-year period doubled the risk for developing hypertension [18]. Many overweight persons do not have hypertension, however, and it is unclear which persons within a population are particularly sensitive [19].
The reason for the association between elevated blood pressure and increased weight is unclear. A possible cause is a decreased renal filtration surface [20], which may lead to renal sodium retention. Obesity also is known to lead to insulin resistance with consequent hyperinsulinemia, and insulin enhances tubular reabsorption of sodium [21]. Enhanced catecholamine activity may also be involved [22], although more definitive data are necessary. Plasma renin activity has also been reported to be elevated in obese persons with hypertension [23]. The increased risk for hypertension also leads to an increased risk for stroke. In the Framingham study, for example, the incidence of stroke went from 22 to 30 to 49 per thousand for men younger than 50 years and from 8 to 14 to 35 per thousand for women in the same age group with relative weights of 110, 110 to 129, and 130 kg or more, respectively. Among persons older than 50 years, these figures went from 70 to 92 to 80 per thousand for men and from 45 to 64 to 121 per thousand for women with relative weights of 110, 110-129, and 130 kg or more, respectively [24].
Blood Lipids
Blood lipid levels are often abnormal in obese persons. High-density lipoprotein (HDL) cholesterol, a higher level of which has been clearly implicated in decreased risk for coronary heart disease, is lower in obese persons [25, 26]. Total and low-density lipoprotein (LDL) cholesterol, however, have been found in cross-sectional studies to be normal [27] or elevated [28] in obese compared with lean persons [28, 29]. In the Framingham study, every 10% increase in relative weight was associated with an increase in plasma cholesterol of 12 mg/dL [30]. Because HDL cholesterol is low and LDL cholesterol is normal to high, the ratio of HDL to LDL cholesterol is generally high, leading to greater atherogenic risk.
The Second National Health Examination Survey (NHANES II) defined hypercholesterolemia as a value at or above 6.47 mmol/L (a clear risk); the relative risk for having such a level was 1.5 times greater in obese than in nonobese persons [12]. In the Tecumseh study, cross-sectional data showed that men younger than 39 years had total cholesterol levels that correlated with relative weight, but no such correlation was found for men older than 40 years [27]. A report of a random sample of Neapolitan men and women found higher total cholesterol levels to be associated with increasing BMI [31]. This effect was more pronounced in men. Triglycerides have generally been found to be higher in obese compared with lean persons [29]. Elevated triglyceride levels have been described with weight gain [32]. Increased free fatty acid availability from enhanced lipolytic activity and hyperinsulinemia enhances the formation of very-low-density lipoprotein (VLDL) in the liver [33]. Also, because lipoprotein lipase activity is decreased, a decreased clearance of triglycerides also occurs [34].
Coronary Heart Disease
Coronary heart disease morbidity is defined generally in epidemiologic studies as nonfatal myocardial infarction and angina pectoris [35]. The role of obesity on coronary heart disease morbidity and mortality has been widely debated. This debate has been fueled by two issues. First, obesity can enhance other risk factors such as high blood pressure, high blood cholesterol and triglyceride levels, low HDL cholesterol, and insulin resistance with hyperinsulinemia that cause coronary heart disease morbidity. Second, obesity has been shown to increase coronary heart disease risk independently [4]. This second point has been more controversial because, as mentioned earlier, a longer period of follow-up in longitudinal studies has been necessary to determine significant associations [46]. Both the Framingham Study [24] and the Los Angeles Heart Study [36] noted a stronger association between body weight and coronary heart disease in their later [12, 13] compared with their earlier reports.
Because most studies have dealt with men, a recent longitudinal study in women has received great interest [35]. This study followed 115 886 U.S. women who were 30 to 55 years old and who were free of coronary heart disease, stroke, and cancer. A higher BMI was positively associated with the occurrence of each category of coronary heart disease. When these categories were combined and adjusted for age and smoking status, relative risk for nonfatal myocardial infarction and fatal coronary heart disease went from 1.0 for a BMI of less than 21, to 1.3 for a BMI of 21 to less than 23, to 1.3 for a BMI of 23 to less than 25, to 1.8 for a BMI of 25 to less than 29, and to 3.3 for a BMI greater than 29 [35]. As expected, control for a history of hypertension, diabetes mellitus, and hypercholesterolemia (conditions known to be biological effects of obesity) attenuated the strength of the association but did not eliminate it. Even mild to moderate overweight increased the risk for coronary heart disease.
Controversy has surrounded the relation of obesity to death. Of seven representative, prospective, high-quality studies of coronary heart disease, four reported a positive association, and three did not [1]. The U.S. Pooling Study compared data from several previously reported longitudinal cohort studies. Again, some showed a positive association, and some did not [22]. When the data from these studies were pooled for greater statistical power, however, a positive relation between obesity and coronary heart disease was noted [37].
Why did the differences in these studies occur? Manson and colleagues [38] recently pointed out that, in 25 major prospective studies on the association between body weight and longevity, each study had at least one of three biases that led to systematic underestimation of the impact of obesity on premature death. These biases included 1) failure to control for cigarette smoking; 2) inappropriate control of the biological effects of obesity, such as hypertension and hyperglycemia; and 3) failure to control for weight loss due to subclinical disease.
Although studies done before the 1980s reported that the lowest mortality rate was found in groups whose body weights were about 15% less than the average weight for height, later studies suggested that, when other factors associated with death were included in the analyses, obesity decreased in importance as an independent risk factor [4]. This decrease occurs because obesity exerts much of its effects through the enhancement of other risk factors such as high blood pressure, diabetes, and elevated lipids [4, 24, 35]. However, if the three biases described previously are recognized and controlled, overall mortality rate is lower for those whose relative weights are somewhat less rather than more than the U.S. average [38].
This argument is convincingly bolstered by the recent study of the relation between BMI and death in male Seventh Day Adventists [39]. This group of nonsmokers was followed for 26 years, and three items are noteworthy. First, no increased mortality rate was noted in the leanest group. The authors state that the relatively large number of subjects who were lean by choice, rather than as a result of preclinical disease or smoking, may explain these findings [39]. This study extends previously reported evidence from the Albany [40] and Kaiser Permanente [41] studies that showed that lean, nonsmoking men do not have an increased risk for death. Second, a significant trend toward increasing mortality rate with increasing BMI for all end points studied (including cancer and cerebrovascular death) was also noted. Thus, a linear, continuous relationship between BMI and death was found. Third, survival analysis was done, and the protective effect associated with low body weight decreased with advancing age and disappeared by 90 years of age. Although the protective effect associated with the lowest quintile of BMI on coronary heart disease death decreased significantly with increasing age, the effect remained greater than one at all ages in men [42].
Diabetes Mellitus
The association between the average weight of population groups and the prevalence of non–insulin-dependent diabetes has been repeatedly observed [43, 44]. The risk for diabetes has been reported to be about twofold in mildly obese, fivefold in moderately obese, and 10-fold in severely obese persons [45]. The duration of obesity is a more important determinant of the risk for developing diabetes [46]. In cross-sectional studies, obesity has been shown to be associated with an increased prevalence of non–insulin-dependent diabetes mellitus in both men and women [47, 48]. The NHANES II data found that the overall relative risk of having diabetes was 2.9 times higher for obese persons who are 20 to 75 years old [12]. The risk for developing diabetes also increases with age [4951], if a family history of diabetes is present [52], and if the obesity is concentrated centrally [53]. A prospective study in Scandinavia showed that moderate obesity was associated with a 10-fold increase in the risk for diabetes [54]. This risk increased steeply as obesity became more severe [54].
Gallbladder Disease
Increasing weight is associated with a greater prevalence of gallbladder disease in both cross-sectional [55] and longitudinal [56] studies. Gallstones occur three or four times more often in obese than in nonobese persons [5557]. The prevalence increases with age and with increasing obesity [58]. Women are particularly at risk, as documented in the longitudinal, prospective Nurses' Health Study [59]. Women with a BMI greater than 30 kg/m2 had a yearly symptomatic gallstone incidence rate of more than 1%, and those with a BMI greater than 45 had a rate of approximately 2% [60]. The gallstones associated with obesity are composed primarily of cholesterol [61]. The mechanism for this increase in gallbladder disease with obesity is due to the cholesterol supersaturation of the bile [62, 63]. In addition, greater gallbladder stasis occurs as obesity supervenes [64].
Respiratory Disease
Obesity affects respiratory function. Increased fat in the chest wall and abdomen reduces lung volume, alters the respiratory pattern, and causes a decreased compliance of the respiratory system [65, 66]. Vital capacity and total lung capacity are frequently diminished. In more severe obesity, a ventilation-perfusion abnormality occurs [67], which is characterized by hypoxia but normal arterial PCO 2 [68, 69]. As the severity of obesity increases, sleep apnea occurs with greater frequency. This condition may be obstructive (due to a combination of excess fatty tissue and increased relaxation of the pharyngeal and glossus muscles) [70], central (due to abnormal control of breathing), or a combination of the two [71]. The full-blown obesity-hypoventilation syndrome is associated with depression of hypercapnic and hypoxic respiratory drives, irregular breathing, frequent apneic periods with resultant severe hypoxia [72], and daytime somnolence. Cor pulmonale may also occur [73].
Cancer
A prospective American Cancer Society study followed 750 000 men and women for 12 years and found that the mortality ratio for cancer for men who were 40% or more overweight was 1.33; the corresponding figure for women was 1.55 [74]. Overweight men had significantly higher mortality ratios for colorectal and prostate cancers, and overweight women had significantly higher rates of endometrial, gallbladder, cervical, ovarian, and breast cancers [74]. In longitudinal, prospective studies, the relation of degree of obesity to increased mortality rate from breast cancer has been well documented in postmenopausal women [7577]. Other studies have generally supported these observations. It has been difficult to differentiate the effect of diet from the effect of obesity. In some types of cancer, such as those of the colon and breast, further study is needed to determine whether obesity simply reflects the diet of the individual patient and whether it is the diet (high fat, high calorie) rather than the degree of obesity that creates the important association.
Gout
The effect of obesity on uric acid levels shows a sexual dimorphism. In the cross-sectional Canadian Health Survey, the percentage of men with uric acid levels greater than 416 mol/L increased from 7% to 31% as the BMI increased from 21 to 31 [78]. Women were not affected until they reached a BMI greater than 31, when the percentage prevalence was 7% [79]. In other cross-sectional surveys, correlations between the level of uric acid and weight repeatedly have been reported [41, 80].
Arthritis
The increased prevalence of osteoarthritis with increasing weight has been described repeatedly in cross-sectional studies [8184]. As weight increased in men and women, the prevalence of osteoarthritis increased from 0.75% to 1.45% in men and from 0.4% to 1.45% in women [84].
Effect of Fat Distribution
The relation between regional fat distribution and health has come to the forefront in the last decade. Several prospective, longitudinal studies have shown a positive association between increased abdominal upper body fat and overall mortality rate [42, 8588]. The issue has been complicated by the fact that these studies have been done by measuring upper body (or central) fat by anthropometric measures (waist-hip or skinfold thickness ratios). It has become clear, however, that visceral fat, rather than central fat, is related to disease risk [89, 90]. Although precise methods for measuring visceral fat exist, they are costly, invasive, and time-consuming.
Central fat distribution is associated with increased risk for coronary heart disease [42, 8588]. The relative risk for death reaches about 2.0 in these studies and is progressive with increasing body fat. Four of the six studies examined men, one examined women, and one examined both sexes. A positive association between central or upper body fat and hypertension has also been described in cross-sectional studies [9193]. Additional cross-sectional studies have shown insulin resistance, hyperinsulinemia, and frank diabetes mellitus to be related to increased abdominal fat [94, 95]. Ohlson and colleagues [96], in an 8-year, prospective, longitudinal study, found an increased risk for developing diabetes over and above the risk for total adiposity per se.
Regional fat distribution has also been related to elevation of serum triglyceride levels and a decrease in HDL cholesterol [97]. In addition, increased abdominal fat has also been related to stroke [98].
Although data on gallbladder disease are still sparse, evidence suggests that abdominal fat distribution enhances the risk [99]. The data for cancer are not as clear, and this topic needs further exploration.
Conclusion
Ample evidence suggests that obesity increases both morbidity and mortality risks and that fat distribution is important as an independent marker of risk. Although controversy surrounds the strength of independent risk compared with the biological effects of obesity on other risk factors, little doubt exists about the severe impact of obesity on health and mortality risks. It is difficult, however, to assign a specific threshold at which health risks begin. The increase in mortality rates in relation to relative weight increase is steeper in men and women younger than 50 years than in older persons, and the increase associated with duration of obesity is also steeper [100]. This finding suggests that a particular effort should be made to prevent weight gain in younger patients. Because little evidence of health risk has been shown at lower weight when adjustments are made for smoking and concurrent disease, it seems reasonable from a public health viewpoint to urge all persons in the United States to maintain an average or somewhat below-average weight. Much can be gained in quality and duration of life and in reduced health costs by such a policy.
- Copyright 2004 by the American College of Physicians
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
