This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

F U T U R E  E L D E R L Y
O B E S I T Y  

26 September 2005  
  

The Health And Cost Consequences 
Of Obesity Among The Future Elderly

An obese seventy-year-old incurs $39,000 in additional
medical costs in old age compared with costs
for a peer who is not obese.

By Darius N. Lakdawalla, Dana P. Goldman, and Baoping Shang

ABSTRACT:

Obesity could have serious consequences for older cohorts. We used a microsimulation to estimate lifetime costs, life expectancy, disease, and disability for seventy-year-olds based on body mass. Obese seventy-year-olds will live about as long as those of normal weight but will spend more than $39,000 more on health care. Moreover, they will enjoy fewer disability-free life years and experience higher rates of diabetes, hypertension, and heart disease. Medicare will spend about 34 percent more on an obese person than on someone of normal weight. Obesity might cost Medicare more than other diseases, because higher costs are not offset by reduced longevity.

Recent evidence suggests that healthy people over age seventy—that is, those with no functional limitations—live longer lives with less lifetime medical spending than their less healthy peers.¹ Moreover, it is well established that risk factors such as obesity, lack of exercise, and an unhealthy diet damage health by increasing the incidence and severity of hypertension and diabetes, and increase medical costs.² This evidence suggests that reducing risk factors could improve health and well-being while still lowering lifetime medical spending, for those covered by private insurance and by third-party payers such as Medicare.

Obesity is a particularly important health risk factor; during the past several decades, it has spread across all age groups and segments of society.³ Fueled by powerful economic forces such as declining food prices and more sedentary work, these trends show no immediate signs of reversal. They could affect the health and medical care demands of the elderly for at least the next several decades.⁴

To understand the consequences of obesity for health and spending, we estimated its effects on the health and medical spending of a cohort of seventy-year-olds. In particular, for the obese, overweight, normal weight, and underweight, we compared expected life span after age seventy, disability-free life span, cumulative health care and Medicare spending at age seventy and beyond, and the prevalence of chronic disease and disability. We focused on the elderly because most public health care spending is devoted to this group.

Obesity seems to be a more pressing public finance problem for the Medicare system than other similar problems, such as smoking, because it lacks clear adverse impacts on longevity but still raises annual health spending. Although the literature finds that obesity is associated with increased mortality among young and middle-aged populations (see, for example, the work of Jay Olshansky and colleagues), it also suggests little or even a negative relationship between obesity and mortality among the elderly.⁵ Our results are consistent with these earlier findings: Obese elderly people live about as long as their nonobese peers, but they experience a poorer quality of life and face higher medical spending. Obesity in one seventy-year-old costs Medicare about $36,000 in expected present value, and it costs other payers about $3,000. Therefore, we estimate that preventing obesity in one person is worth at least $39,000 in gross value to society, in 2004 dollars. That is, if such an outcome could be achieved at a cost under $39,000, it would improve welfare.

Study Data And Methods

Data source. Data for this study come from the Medicare Current Beneficiary Survey (MCBS), a nationally representative sample of the entire Medicare population, both institutionalized and noninstitutionalized. The sample frame consists of aged and disabled beneficiaries enrolled in Medicare Part A or Part B, or both. The disabled (under age sixty-five) and the oldest old (over age eighty-five) are oversampled. The first round of interviewing in this annual panel survey was conducted in 1991. The MCBS contains administrative data on Medicare spending along with self-reported data on non-Medicare health spending. Each sample member’s data are weighted to be representative of the Medicare population.

Body mass index. We used self-reported height and weight in the MCBS to compute body mass index (BMI: kg/m²) and classified individuals into four weight classes: underweight (BMI less than 20), normal weight (BMI 20–24.9), overweight (BMI 25–29.9), and obese (BMI 30 or more). Since previous research has documented systematic biases in these measures, we used the correction for self-reported weight proposed by John Cawley.⁶ This correction had little impact on our results, primarily because the bias in BMI is too small to affect a person’s weight class.

Disease and disability. We chose health status measures that were predictive of costs and available in the MCBS. For a variety of diseases, the MCBS asks whether the respondent has ever been diagnosed by a doctor with that condition. For each disease, we estimated the proportion of life years a person had that condition and then constructed the average proportion within each weight group. We focused on diseases for which self-reports were available: diabetes, cancer (excluding skin), heart disease (myocardial infarction, angina, coronary heart disease, congestive heart failure, valve disease, arrythmias, or other heart problems), hypertension, Alzheimer’s disease, stroke, lung disease (emphysema, asthma, or chronic obstructive pulmonary disease), and osteoarthritis. The survey also asks respondents about limitations in their activities of daily living (ADLs): bathing, dressing, eating, getting into and out of a chair, walking, and using the toilet. We constructed four classes of disability: no ADL limitations, one or two ADL limitations, at least three ADL limitations, and institutionalized. These four categories are mutually exclusive, so that people in any of the first three categories are by definition noninstitutionalized. It was necessary to separate institutionalized beneficiaries from all others because from 1996 onward, MCBS respondents (or their proxies) were asked different disability questions depending on whether or not they lived in nursing homes.

Exhibit 1 reports the baseline characteristics of the seventy-year-old cohort in the MCBS. The MCBS was designed to be nationally representative of the elderly population. Its subsamples of people ages 65–69 and age 70 or older are also nationally representative of those age groups. Although it is not necessarily representative for people at age seventy, we have found very similar results for different cohorts; thus, we do not believe that idiosyncratic sample selection is responsible for our results.

Mean medical spending is higher for obese and overweight seventy-year-olds than for all other groups. The differences for the cohort of seventy-year-olds is not statistically significant, but this seems to be a matter of sample size: Computing the same statistics for MCBS enrollees ages 70–80, for example, yields qualitatively similar but statistically distinguishable results. Obese and underweight seventy-year-olds have the statistically highest prevalence of disability, distinguishable from all the other groups. The result for the underweight is likely attributable to the fact that illness can contribute to weight loss. Diabetes, heart disease, hypertension, and osteoarthritis are substantially and statistically higher at baseline for the obese and overweight, compared with those of normal weight. For these conditions, the underweight are no worse off or slightly better off than those of normal weight.

Simulation. To simulate the consequences of obesity, we used the Future Elderly Model (FEM) to track the health conditions, functional status, and Medicare and total health care spending for obese and nonobese beneficiaries. Obesity status was measured at age seventy; thus, we simulated the old-age health costs for obese and nonobese seventy-year-olds. The FEM consists of three components: a model of health care costs; a model of health status transitions; and a model to predict characteristics of future, newly entering Medicare enrollees.
The latter component was not needed for this paper since we followed a single representative cohort of seventy-year-olds in 1998. Our simulation starts in 1998 and follows the entire cohort—tracking their costs, health, and disability—until everyone in the cohort has died. More detail on how we computed disease, disability, and costs is provided in a RAND technical report.⁷ The standard errors of our estimates were computed via bootstrap replications: We sampled the original cohort of seventy-year-olds (with replacement) 1,000 times and computed the experiences of each sampled cohort; standard errors were computed across the results for these 1,000 simulations. These yielded standard errors of the simulation; that is, they took the underlying microsimulation model as given and computed variations that arose because of differences in the population of respondents and the randomness in the model’s transitions.

Health status transitions. The FEM first predicts the health conditions and functional status of the baseline sample for the next year. To simulate health transitions, a discrete piecewise linear hazard model was first estimated. The hazard of getting a disease and dying depends on risk factors (sex, education, race, ethnicity, obesity, underweight, ever having smoked); other health conditions; functional status; and age (piecewise linear spline, node at age seventy-seven). Similar models were used to predict functional status and institutionalization status. We treated all health states as “absorbing”—that is, once someone gets an illness, he or she has it forever and thus cannot get it again—and modeled transitions into the states. This assumption was consistent with the way the data were obtained (“Has a doctor ever told you…”) and with the course of most of the chronic diseases. However, for some conditions or outcomes, such as altered functional status, recovery is possible; therefore, the hazard model would overestimate their prevalence. We simulated health transitions by first predicting each person’s hazard of dying, getting a new health condition, or achieving a new functional status. We then drew random numbers and compared them with the predicted hazard rate to determine the simulated outcome.

Expenditures. Expenditures were based on pooled weighted least squares regressions, with total Medicare reimbursement and total health care reimbursement as the dependent variables, and health status measures, self-reported disease categories, and interactions of health measures and disease conditions as the independent variables. The model was calibrated to replicate total health care and Medicare spending for the elderly sample represented by the MCBS. All FEM costs are in 1998 dollars and are adjusted for inflation using the medical Consumer Price Index (CPI).

We dropped Medicare health maintenance organization (HMO) enrollees and assumed that all Medicare beneficiaries were covered under fee-for-service (FFS) Medicare. This might have caused us to understate the relationship between obesity and Medicare outlays. To see why, note that Medicare payments to HMOs do not depend on obesity status, but actual HMO expenditures might. As a result, Medicare actually saves money on obese HMO enrollees, while HMOs themselves lose money on patients who are more costly than average.

For each person in the cohort, our simulation yielded a complete old-age profile of medical costs, a disease history, and an age at death. Based on these profiles, we estimated, for each of the four weight classes, the following three sets of health indicators: expected healthy and frail life span, old-age disease prevalence, and old-age medical spending.

For spending, we estimated average annual Medicare spending and average annual health care spending in total. Old-age expenditures are the sum of expenditures over each person’s simulated remaining lifetime and should be interpreted as the present value of old-age expenditures at age seventy. This follows the approach of James Lubitz and colleagues.⁸ Implicitly, we assumed that the rate of growth in real Medicare spending is equal to the real risk-free rate of discount, which is usually taken to be at or slightly above 3 percent.⁹ From Medicare’s inception, its real per capita spending has grown at around 4 percent per year, although growth over the past ten years has been closer to 2 percent.¹⁰

Longevity. We estimated both total life span and life span spent in each disability state. Expected life span is equal to average lifetime within each group. Similarly, we calculated expected years spent in each disability state—no ADL limitations, one or two ADL limitations, and at least three ADL limitations—and estimated the average years in each disability state for each weight class. To estimate the impact of each disease (including institutionalization status), for each person, we estimated the proportion of life years spent with the conditions, and we estimated the average proportion within each weight class.

Study Results

We obtained three sets of health indicators for each weight class in our sample. The first, related to longevity and disability-free life expectancy, is shown in Exhibit 2. The total height of each bar is equal to total life expectancy within the group. Most of the estimates are statistically different from all other groups. Nearly all estimates for the obese are statistically different from all other weight groups, with just two exceptions: Institutionalized years are statistically indistinguishable for the obese and the overweight, as are years spent with one or two ADLs. Total life expectancy is highest for the overweight, followed by the obese, the normal weight, and the underweight.

Exhibit 2 also shows the effects of body mass on active life expectancy. The obese can expect 4.0 disability-free life years—the lowest disability-free life expectancy of any group. The underweight are next at 5.8 disability-free years. The overweight live for 6.6 disability-free years, while the normal-weight group lives without disability for an average of 6.8 years. Obese seventy-year-olds can expect to spend 40 percent more time in disability than their normal-weight counterparts. The obese also spend about 40 percent more time in the two most disabled states (at least three ADLs or institutionalized) than any other group. They will be in one of these two states for an average of 4.7 years, while the next most disabled group, the underweight, spends just 3.3 years in these states on average. Our estimated longevity of 14.0 years is very similar to conventional life-table estimates for this cohort, which range from 14.26 to 14.30.¹¹ Our estimate of 6.2 years of overall disability-free life expectancy is similar to the estimate of Lubitz and colleagues, who estimate that at age seventy, there are 6.9 expected active life years—which they define as years without any limitation or with only a Nagi limitation (difficulty stooping, crouching, lifting objects up to ten pounds, extending the arms above the shoulder, grasping small objects, or walking two to three blocks).¹²

Exhibit 3 shows that higher disability largely translates into higher health spending. The obese and the underweight are the highest-spending groups in annual terms. The obese spend about $17,500 annually, of which about $11,500 is Medicare’s responsibility. Similarly, the underweight spend just over $17,600, of which about $10,200 is Medicare’s responsibility. For these two groups, total annual expenditures are more than 15 percent higher than for the normal-weight and overweight groups; the two groups are statistically identical to each other but different from all the other groups. However, over their remaining lifetimes, the obese are the most costly. From age seventy, an obese person can expect to spend $230,000 on health care in present-value terms, $143,000 of which is borne by Medicare. This is the highest spending level for any group and is statistically different from all of the other groups. The total burden on Medicare is more than 20 percent higher than for the next closest group (the overweight). The obese consistently spend more per year than any other group except for the underweight, but they face these high expenses for more years than the underweight do.

A more direct measure of old-age health for each weight group is provided in Exhibit 4. The letters on each bar denote statistical differences across groups. We focused on four diseases with the largest gradient across weight classes: diabetes, heart disease, high blood pressure, and osteoarthritis. We also included two diseases that are particularly prevalent for the underweight: cancer and lung disease. The largest absolute gradient across weight classes can be found for diabetes: Obese people suffer from diabetes for 41 percent of their life years, while normal-weight people suffer for just 17 percent of their life years. The gradient in hypertension is also quite large. All of these differences are statistically significant and refer to the remaining lifetimes of people age seventy and older. In contrast, there are virtually no differences in cancer or lung disease across weight classes.

Discussion And Policy Implications

From age seventy onward, Medicare spends 35 percent more on an obese person than on someone of normal weight. Moreover, people who are obese at age seventy experience far fewer disability-free life years and considerably higher prevalence of disease than their normal-weight peers. Therefore, from the point of view of the Medicare system and the rest of the publicly financed health care system, the growth in obesity could entail considerable financial costs. Our estimates suggest that the expected cost to Medicare of obesity in one seventy-year-old beneficiary is about $36,000 over the person’s remaining lifetime. There is an additional $3,000 cost to other payers (the individual, Medicaid, private insurers, and other non-Medicare sources).

Our cost estimates are similar to others in the literature. Martha Daviglus and colleagues have studied the relationship of BMI in young adulthood and middle age and future Medicare spending.¹³ Our results are qualitatively similar to their finding that obesity is linked to higher annual and lifetime spending. Eric Finkelstein and colleagues have estimated that obesity increased Medicare enrollees’ medical spending by an average of $1,486 annually in 1998, or $1,903 in 2004 dollars.¹⁴ This is statistically similar to our estimate, $2,325 (with a standard error of $318).

Lubitz and colleagues have produced another important and relevant set of results. They found that old-age spending is lower for healthy seventy-year-olds than for their institutionalized peers, in spite of the longer life expectancy of the healthy.¹⁵ However, old-age expenditures were fairly similar across other disability groups. Across disability groups, therefore, Lubitz and colleagues found that old-age spending varies little, as the longer life expectancies of healthier groups are completely offset by lower annual medical spending. Our results differ because we do not find substantially lower life expectancy for the obese, in spite of their higher medical spending.¹⁶ This fits with the existing literature, which consistently finds a relationship between obesity and medical spending but little (or perhaps even a negative) correlation between obesity and mortality among the elderly.

The literature finds that underweight people have lower expected life spans, but there is no consensus on mortality differentials between the normal-weight and the overweight or obese.¹⁷ Some have argued that overweight and obesity can be protective for the elderly, while others find that it increases mortality just as it does for the young. The different relationships for elderly and young populations could have clinical or biological causes, but they could also result from mortality selection. That is, obese people who survive into old age might be particularly robust; this could offset a (hypothetically) negative causal effect of obesity on longevity.

Our analysis suggests that the focus on and controversy concerning total life expectancy might sidestep the most important issue. Unlike the difference in total life expectancy—which by our calculation and by the lack of consensus in the literature seems somewhat small—the differences in disability-free life expectancy are stark and in line with the well-known relationship between obesity and diseases such as hypertension, diabetes, and heart disease. Obese seventy-year-olds can expect to face 2.8 additional years of ADL disability, two years of which will involve moderate to severe levels of disability (three or more ADL limitations or institutionalization).

These numbers suggest that the disability effects of obesity, rather than the effects on medical spending, might be the larger component of obesity’s welfare costs. The U.S. Centers for Disease Control and Prevention (CDC) has suggested that a year of life spent with an ADL limitation is about half as valuable as a year of healthy life.¹⁸ Therefore, the cost of 2.8 additional disabled life years is similar to losing 1.4 years of life. Since a year of life to an elderly person is easily worth more than $28,000, the cost of disability must be larger than $39,000.¹⁹

Although average population weights have likely been rising over the past several centuries, the past several decades have seen particularly rapid growth.²⁰ The rate of obesity has more than doubled, affecting every age group. Close to half of the U.S. population is overweight, and more Americans are obese than smoke, use illegal drugs, or suffer from ailments unrelated to obesity.²¹ There is some optimism that these trends could be reversed. James Hill and colleagues estimate that modest changes in caloric intake or increases in physical activity—about 100 kilocalories per day, equivalent to a fifteen-minute walk—would be enough.²² Samuel Preston has echoed this sentiment.²³ But changing behavior is notoriously difficult. The recent acceleration in obesity has likely been stimulated by rapid reductions in food prices and innovations in food production and processing, none of which is likely to disappear.²⁴

Ultimately, the impact of obesity treatment on medical costs and health outcomes depends on the extent to which obesity is the primary driver of health differences between the obese and nonobese. If the relationship is causal, preventing or curing obesity in one person will restore that person to the spending level of someone of normal weight. More clinical and epidemiological research is needed to establish the precise extent of causality, but existing evidence points toward a causal relationship.

In clinical trials and observational studies, weight loss is consistently related to improvements in health.²⁵ The fact that these improvements occur for a variety of very different weight-loss mechanisms suggests that weight loss itself improves health, instead of the possibility that a particular feature of a weight-loss device (for example, a particular chemical byproduct of pharmaceutical therapy) is responsible for health improvement.

Randomized clinical trials of behavioral changes have suggested that weight loss induced by diet and exercise reduces the risk of diabetes and hypertension and improves health.²⁶ Similar findings have been found in trials of pharmaceutically induced weight loss. The weight-loss drug sibutramine, for example, has been found both to lower weight and to reduce hypertension and diabetes prevalence.²⁷

Our results, coupled with the literature on weight loss and health improvements, point to the importance of evaluating the health and cost consequences of weight loss in clinical trials. Such trials represent the most scientifically rigorous way of establishing and quantifying the degree of causality. They would also provide insight into the types of treatments that could save Medicare money and safeguard the welfare of Medicare enrollees.

A complete policy approach to the problem of obesity considers the public costs of obesity to government health insurance schemes, the financial costs of treating the disease, and the private nonfinancial costs of treatment. As examples of the latter, people may dislike exercise, like fatty or unhealthy food, dislike surgery, and so on. Our analysis examines one component of this overall picture: the relationship between obesity and health care spending, both public and private.

Principal funding for this study came from the Centers for Medicare and Medicaid Services (CMS Contract no. 500-95-0056), with additional funding from the National Institute on Aging through its support of the RAND Roybal Center for Health Policy Simulation (P30AG024968) and the RAND Center for the Study of Aging (P30AG12815). The authors are solely responsible for the paper’s contents. No statement in this paper should be construed as being an official position of the CMS.

NOTES

1. J. Lubitz et al., “Health, Life Expectancy, and Health Care Spending among the Elderly,” New England Journal of Medicine 349, no. 11 (2003): 1048–1055.
2. L.J. Appel et al., “Effects of Comprehensive Lifestyle Modification on Blood Pressure Control: Main Results of the PREMIER Clinical Trial,” Journal of the American Medical Association 289, no. 16 (2003): 2083–2093; H. Gin, V. Rigalleau, and L. Baillet, “Diet and Physical Activity in Type 2 Diabetes Prevention” (original in French), Revue du praticien 53, no. 10 (2003): 1074–1077; R. Sturm, “The Effects of Obesity, Smoking, and Drinking on Medical Problems and Costs,” Health Affairs 21, no. 2 (2002): 245–253; and D. Thompson et al., “Body Mass Index and Future Healthcare Costs: A Retrospective Cohort Study,” Obesity Research 9, no. 3 (2001): 210–218.
3. K.M. Flegal et al., “Overweight and Obesity in the United States: Prevalence and Trends, 1960–1994,” International Journal of Obesity and Related Metabolic Disorders 22, no. 1 (1998): 39–47; and K.M. Flegal et al., “Prevalence and Trends in Obesity among U.S. Adults, 1999–2000,” Journal of the American Medical Association 288, no. 14 (2002): 1723–1727.
4. D.M. Cutler, E.L. Glaeser, and J.M. Shapiro, “Why Have Americans Become More Obese?” Journal of Economic Perspectives 17, no. 3 (2003): 93–118; S.-Y. Chou, M. Grossman, and H. Saffer, “An Economic Analysis of Adult Obesity: Results from the Behavioral Risk Factor Surveillance System,” Journal of Health Economics 23, no. 3 (2004): 565–587; and D.N. Lakdawalla and T. J. Philipson, “Technological Change and the Growth of Obesity,” NBER Working Paper no. 8946 (Cambridge, Mass.: National Bureau of Economic Research, 2002).
5. S.J. Olshansky et al., “A Potential Decline in Life Expectancy in the United States in the Twenty-first Century,” New England Journal of Medicine 352, no. 11 (2005): 1138–1145.
6. J. Stevens et al., “Accuracy of Current, Four-Year, and Twenty-Eight-Year Self-Reported Body Weight in an Elderly Population,” American Journal of Epidemiology 132, no. 6 (1990): 1156–1163; A.J. Stunkard and J.M. Albaum, “The Accuracy of Self-Reported Weights,” American Journal of Clinical Nutrition 34, no. 8 (1981): 1593–1599; and J. Cawley, “Impact of Obesity on Wages,” Journal of Human Resources 39, no. 2 (2004): 451–474.
7. D. Goldman et al., Health Status and Medical Treatment of the Future Elderly: Final Report, Pub. no. TR-169 (Santa Monica, Calif.: RAND, 2004). More details are available in an online technical appendix,
content.healthaffairs.org/cgi/content/full/hlthaff.w5.r30/DC2.
8. Lubitz et al., “Health, Life Expectancy, and Health Care Spending.”
9. J.J. Siegel, “The Real Rate of Interest from 1800–1990: A Study of the U.S. and the U.K,” Journal of Monetary Economics 29, no. 2 (1992): 227–252.
10. J. Bhattacharya and D.N. Lakdawalla, “Does Medicare Benefit the Poor?” Journal of Public Economics (forthcoming).
11. Data are from the Human Mortality Database, University of California, Berkeley, and Max Planck Institute for Demographic Research, available at www.mortality.org.
12. Lubitz et al., “Health, Life Expectancy, and Health Care Spending.”
13. M.L. Daviglus et al., “Relation of Body Mass Index in Young Adulthood and Middle Age to Medicare Expenditures in Older Age,” Journal of the American Medical Association 292, no. 22 (2004): 2743–2749.
14. E.A. Finkelstein, I.C. Fiebelkorn and G. Wang, “National Medical Spending Attributable to Overweight and Obesity: How Much, and Who’s Paying?” Health Affairs, 14 May 2003,
content.healthaffairs.org/cgi/content/abstract/hlthaff.w3.219 (22 July 2005).
15. Lubitz et al., “Health, Life Expectancy, and Health Care Spending.”
16. Ibid.
17. D.C. Grabowski and J.E. Ellis, “High Body Mass Index Does Not Predict Mortality in Older People: Analysis of the Longitudinal Study of Aging,” Journal of the American Geriatrics Society 49, no. 7 (2001): 968–979; and E.E. Calle et al., “Body-Mass Index and Mortality in a Prospective Cohort of U.S. Adults,” New England Journal of Medicine 341, no. 15 (1999): 1097–1105.
18. P. Erickson, R. Wilson, and I. Shannon, “Years of Healthy Life,” Healthy People 2000 Statistical Notes no. 7 (1995): 1–15.
19. J.E. Aldy and W.K. Viscusi, “Age Variations in Workers’ Value of Statistical Life,” NBER Working Paper no. 10199 (Cambridge, Mass.: NBER, 2004).
20. D. Costa and R. Steckel, “Long-Term Trends in Health, Welfare, and Economic Growth in the United States,” NBER Working Paper no. 76 (Cambridge, Mass.: NBER, 1995).
21. A.M. Wolf and G.A. Colditz, “Current Estimates of the Economic Cost of Obesity in the United States,” Obesity Research 6, no. 2 (1998): 97–106; and J. Tuomilehto et al., “Prevention of Type 2 Diabetes Mellitus by Changes in Lifestyle among Subjects with Impaired Glucose Tolerance,” New England Journal of Medicine 344, no. 18 (2001): 1343–1350.
22. J.O. Hill et al., “Obesity and the Environment: Where Do We Go from Here?” Science 299, no. 5608 (2003): 853–855.
23. S.H. Preston, “Deadweight? The Influence of Obesity on Longevity,” New England Journal of Medicine 352, no. 11 (2005): 1135–1137.
24. Cutler et al.,“Why Have Americans Become More Obese?”; Chou et al., “An Economic Analysis of Adult Obesity”; and Lakdawalla and Philipson, “Technological Change and the Growth of Obesity.”
25. E.R. Pamuk et al., “Weight Loss and Mortality in a National Cohort of Adults, 1971–1987,” American Journal of Epidemiology 136, no. 6 (1992): 686–697; and E.R. Pamuk et al., “Weight Loss and Subsequent Death in a Cohort of U.S. Adults,” Annals of Internal Medicine 119, no. 7, Part 2 (1993): 744–748.
26. G.A. Bray, D.H. Ryan, and D.W. Harsha, “Diet, Weight Loss, and Cardiovascular Disease Prevention,” Current Treatment Options in Cardiovascular Medicine 5, no. 4 (2003): 259–269.
27. W.P. James et al., “Effect of Sibutramine on Weight Maintenance after Weight Loss: A Randomised Trial,” Lancet 356, no. 9248 (2000): 2119–2125; and S. O’Meara et al., “The Clinical Effectiveness and Cost-Effectiveness of Sibutramine in the Management of Obesity: A Technology Assessment,” Health Technology Assessment 6, no. 6 (2002): 1–97.

Darius Lakdawalla is an economist at RAND in Santa Monica, California. Dana Goldman (dgoldman{at}rand.org) is corporate chair and director of health economics there. Baoping Shang is a fellow at the Pardee RAND Graduate School.

Access the table of contents for this package

DOI: 10.1377/hlthaff.W5.R30
©2005 Project HOPE–The People-to-People Health Foundation, Inc.