Objectives. We estimated health care expenditures associated with overweight and obesity and examined the influence of age, race, and gender.

Methods. Using 1998 Medical Expenditure Panel Survey data, we employed 2-stage modeling to estimate annual health care expenditures associated with high body mass index (BMI) and examine interactions between demographic factors and BMI.

Results. Overall, the mean per capita annual health care expenditure (converted to December 2003 dollars) was $3338 before adjustment. While the adjusted expenditure was $2127 (90% confidence interval [CI]=$1927, $2362) for a typical normal-weight White woman aged 35 to 44 years, expenditures were $2358 (90% CI=$2128, $2604) for women with BMIs of 25 to 29.9 kg/m2, $2873 (90% CI=$2530, $3236) for women with BMIs of 30 to 34.9 kg/m2, $3058 (90% CI=$2529, $3630) for women with BMIs of 35 to 39.9 kg/m2, and $3506 (90% CI=$2912, $4228) for women with BMIs of 40 kg/m2 or higher. Expenditures related to higher BMI rose dramatically among White and older adults but not among Blacks or those younger than 35 years. We found no interaction between BMI and gender.

Conclusions. Health care costs associated with overweight and obesity are substantial and vary according to race and age.

Obesity is the second-leading cause of preventable death in the United States, and it is a major cause of morbidity and disability.1,2 Over the past 2 decades, the prevalence of obesity has risen substantially; 64% of Americans are now overweight, and 30% are obese (i.e., have a body mass index [BMI] in excess of 30 kg/m2).3 This increase has led to concerns about a possible epidemic of adverse health outcomes and their associated economic consequences.

Previous researchers have examined the effect of obesity on total health care costs using different methodologies.4–8 Some have determined the costs attributable to obesity by estimating people’s likelihood of developing and incurring costs associated with weight-related diseases such as hypertension, diabetes, and heart disease.4,6 Others have used cross-sectional data at the individual level to examine health care expenditures among people at different BMI levels.5–8 While estimates vary depending on the methodology used and when the study was conducted, the most recent estimates place the excess costs attributable to overweight and obesity at between 4% and 9% of total health care costs.4–8 In addition, some studies have shown that these increased costs are associated with more use of health care resources.5

Few researchers examining the health care costs of obesity, however, consider whether patient demographic factors affect the relationship between BMI and cost. It is generally accepted that there is a dose–response relationship between higher BMIs and increased adverse health consequences, including mortality,1,2,9 but whether this relationship exists for certain race and age subgroups is more controversial. For instance, some studies suggest that the higher risk of death attributable to obesity is blunted—and, in some instances, nonexistent—among African Americans and older adults.10–12 Differences in obesity-related disease burden in different demographic populations may lead to differences in obesity-related health care costs. At least one study, involving data gathered in 1987, revealed that health care costs associated with higher BMI rose more steeply among women than men among those with BMIs at above-normal levels.13

Understanding the influence of demographic factors on the costs attributable to obesity can facilitate accurate projections of future health care costs given that current increases in obesity prevalence rates in the United States disproportionately affect women, Hispanics, and Blacks.3,14 In our study, we examined annual per capita health care expenditures associated with overweight and obesity among US adults and assessed the influences of gender, age, and race/ethnicity on this relationship.

Data Source

The Medical Expenditure Panel Survey (MEPS), a national longitudinal survey funded by the Agency for Healthcare Research and Quality, collects nationally representative information on health service use including expenditures, types and settings of care, payment and insurance information, and patient health status and diagnoses (for details, see http://www.meps.ahrq.gov/).

Respondents provided information on health care use by completing a calendar and were instructed to keep track of all health care expenditures, including bills and receipts for care. They also provided sociodemographic, insurance, and employment information. Some expenditure data were verified through records of medical providers, health insurers, and pharmacies, which were believed to provide more accurate information than self-reports.

In instances in which medical expenditures were unclear for a particular medical event, an imputation procedure was performed. In general, this procedure imputed data from events for which complete information was available to events in which information was missing but characteristics were similar (e.g., charge information, demographic factors, provider type, and characteristics of the medical event). This procedure was used, for example, to provide estimates for care delivered under capitated reimbursement arrangements or through public clinics or veterans’ hospitals and to adjust for household-reported insurance payments, given that respondents were often unaware that their insurer had paid a discounted amount to the provider.

Before public release, data were edited to correct for several problems, including outliers, copayments or charges reported as total payments, and reimbursed amounts reported as out-of-pocket payments. Health expenditures included copayments and other out-of-pocket expenses incurred by respondents but did not include expenditures for over-the-counter medications or alternative care services.

The MEPS sample was derived from a subsample of individuals who took part in another national survey, the National Health Interview Survey (NHIS), conducted by the National Center for Health Statistics (for details, see http://www.cdc.gov/nchs/nhis.htm). When appropriately weighted, results from the NHIS can be generalized to the noninstitutionalized, civilian population of the United States (the sampling frame specifically excludes institutionalized individuals, prisoners, and military personnel). Our study used data from the 1998 MEPS sample, which included households representing subsamples of the 1996 and 1997 versions of the NHIS. The combined response rate was 65%.

Although weight and height data were not collected as part of MEPS, these measurements were ascertained from all 1996 NHIS respondents 18 years or older. However, these data were available for only one adult per household in the 1997 NHIS.15 Hence, a substantial subsample of MEPS respondents (27%) did not have data available on height or weight and were not included in our study. Our sample differed slightly from the ineligible MEPS sample in that the present respondents were significantly more likely to be White (76% vs 73%) and female (53% vs 49%) and to have accrued health care expenditures (84% vs 79%). Also, our sample was older on average (45 vs 42 years).


We calculated BMI by dividing weight in kilograms by the square of height in meters. We then grouped respondents, according to their BMI, as underweight (less than 18.5 kg/m2), normal weight (18.5–24.9 kg/m2), overweight (25.0–29.9 kg/m2), obese class I (30.0–34.9 kg/m2), obese class II (35.0–39.9 kg/m2), or obese class III (≥ 40.0 kg/m2) in keeping with national guidelines.16

In MEPS, expenditures were classified as those incurred in areas such as inpatient care, outpatient hospital-based care (e.g., out-patient procedures), ambulatory visits, emergency room visits, and prescription medications. Total annual expenditures represented those accrued through conventional care, including “out-of-pocket” expenditures made by respondents.

In addition, we extracted several factors we believed to be potential confounders of the relationship between BMI and total annual expenditures but not part of the causal pathway. These factors included age, gender, race/ethnicity (White, Black, Hispanic, other), educational level (less than high school, high school, college), income status, region of the country (Northeast, South, Midwest, West), and whether the respondent resided in a metropolitan or rural setting. We also examined type of insurance coverage (private, Medicare, other public, uninsured). Data on smoking status (former, never, current) were available only for 1997 NHIS respondents.

Data Analysis

We conducted multivariable analyses to estimate annual health care expenditures among adults in the 6 BMI categories, using normal-weight individuals as the reference group. Because a substantial minority of respondents had accrued no expenditures, we used the 2-stage modeling approach of Duan et al. to estimate average expenditures in the overall population.17 According to this approach, expenditures are modeled as a product of probability of health care use and predicted expenditure level conditional on the presence of expenditures.

In the first stage, we used logistic regression to model the probability of the presence of expenditures. In the second stage, we used linear regression to model the relationship between BMI and the logarithm of annual health care expenditures among those accruing expenditures. Both models were adjusted for age (in decades), gender, race, education, type of insurance coverage, region of the country, and rural versus metropolitan residence. Additional adjustments for income (greater or less than $25 000) in the overall sample and for smoking history in the sub-sample with complete data did not change our results substantially, and these data are not described here.

To guard against departures from normality on the logarithmic scale and to account for the complex survey design, we used a weighted version of the Duan et al. smearing estimate to arrive at annual per capita expenditures on the original scale of the 2-stage model.17 We modified a bootstrapping approach18 to arrive at 90% confidence intervals (CIs) for our estimates that accounted for the weighting procedures and complex survey design. We selected 1000 bootstrap samples via stratified sampling, with replacement within each stratum and primary sampling unit combination of a size equal to the original frequency. We converted all expenditures to December 2003 dollars using the Medical Consumer Price Index (which had shown an increase of 24.5% since July 1998).19 Results are presented for typical men and women with the most prevalent sociodemographic characteristics.

To characterize the relationship between BMI and health care expenditures more fully, we conducted several additional analyses. We used the 2-stage modeling approach to evaluate the relationship between BMI and expenditures for major categories of care such as inpatient care, outpatient hospital-based care, ambulatory and emergency room visits, and medications; these categories were responsible for the bulk of total annual health care expenditures.

To examine whether demographic factors influenced the relationship between BMI and total health care expenditures, we examined the interaction between BMI and 3 factors: gender, race/ethnicity (White, Black, or Hispanic), and age group (less than 35, 35–54, 55–64, or 65 years or above). Because of the limited numbers available in some of the categories, we combined certain categories, allowing us to estimate 90% confidence intervals using bootstrapping. In the race/ethnicity–BMI analyses, we treated individuals with BMIs of 35 kg/m2 and higher as a single group. In the age–BMI analyses, we excluded underweight respondents and categorized respondents with BMIs of 30 or higher into one group. These latter analyses were consistent with interaction models in which all 6 BMI groups were specified.

Given the complex sampling design, we conducted all analyses using SAS version 8 (SAS Institute Inc, Cary, NC) and SUDAAN version 8.00 (Research Triangle Institute, Research Triangle Park, NC) to obtain valid standard errors.20,21 We weighted samples to reflect population estimates.

BMI and Health Care Expenditures and Use

Of the 11 212 adults 18 years or older in our study sample, 53% were women; 76% were White, 12% were Black, and 10% were Hispanic. In addition, 32% were younger than 35 years, 40% were aged 35 to 54 years, 11% were aged 55 to 64 years, and 17% were 65 years or older. Finally, 44% of the respondents were in the normal weight range, 35% were overweight, and 17% were obese.

Eighty-five percent of the respondents reported health care expenditures in the preceding year. Table 1 displays weighted median per person annual totals for overall health care expenditures and for the 5 major categories of expenditures before adjustment. The weighted mean annual health care expenditure (in December 2003 dollars) was $3338 per person before adjustment. In comparison with a mean expenditure of $2970 among normal-weight adults, mean expenditures were $3038 for overweight adults and $4333 for obese adults.

After adjustment for type of insurance coverage and sociodemographic and geographic factors, respondents at higher BMI levels were more likely than normal-weight respondents to have accrued expenditures (P < .001 for model stage 1); moreover, among those who accrued expenditures, higher BMI was associated with higher expenditures (P < .001 for model stage 2). Figure 1 illustrates adjusted annual expenditures, according to BMI, for a typical 35- to 44-year-old White man and woman with the most prevalent sociodemographic characteristics observed in the sample: high school education, private insurance coverage, and residence in a southern metropolitan setting.

For most major types of health care services, we found higher expenditures among those with higher BMIs (Figure 1). Obese adults had significantly higher medication and office visit expenditures (all Ps < .001). Prescription medication expenditures exhibited the highest relative increases associated with BMIs above normal. Higher BMI was also associated with inpatient (P < .001) and emergency room (P < .006) expenditures, but they did not predict expenditure levels among those who had accrued expenditures.

Effects of Demographic Factors on the BMI–Expenditures Relationship

Mean total annual health care expenditures were $2703 for men and $3895 for women; in addition, mean expenditures were $3610 for Whites, $3014 for Blacks, and $2107 for Hispanics. The rise in expenditure associated with overweight and obesity was similar among men and women but differed significantly according to race/ethnicity. Overall unadjusted mean expenditures were lower for Black than for White adults; however, the mean expenditure among normal-weight Black adults was higher ($3491) than that among normal-weight white adults ($3068). The mean expenditure among normal-weight Hispanic adults was $2346.

The rise in expenditure associated with higher BMI was most dramatic for Whites after adjustment for other factors (Figure 2). Among Hispanics, the association was attenuated, and expenditures did not appear to increase until BMI exceeded 30 kg/m2; however, confidence intervals were wide. The relationship among Black adults was inconsistent. The interaction between BMI and race was statistically significant for stage 2 of the model, but not stage 1, suggesting that the relationship between higher BMI and the probability of accruing expenditures did not differ according to race/ethnicity; however, predicted BMI-specific expenditures among those who had accrued expenditures differed significantly according to race.

In post hoc analyses, we examined the associations between high BMI and the 3 largest categories of expenditures for White, Black, and Hispanic adults separately. Among Whites, higher BMI was associated with higher expenditures for inpatient stays, office visits, and medications. Among Blacks, BMI was not associated with likelihood of inpatient expenditures, but in the case of the 9.4% of Black respondents who had accrued inpatient expenditures, higher BMI was associated with lower expenditures (P = .005). However, Black adults with higher BMIs were more likely to report office visit and medication expenditures. Among Hispanic respondents, higher BMIs were not associated with inpatient expenditures but were associated with expenditures for office visits and medications (both Ps < .001).

Annual health care expenditures rose in a stepwise fashion with higher BMI for all age groups other than those younger than 35 years. The relative rise was more substantial as adults increased in age (Figure 3), with the largest increase observed among those 55 years or older. In addition, although age did not influence the relationship between BMI and accrual of health care expenditures (stage 1 of model), it did exhibit an effect on the relationship between BMI and total expenditures among those who had accrued expenditures (P < .001).

We found the excess health care expenditures associated with overweight and obesity to be substantial. Higher expenditures attributable to obesity were seen for all major forms of care examined. The relative increase in weight-related health care expenditures was similar for men and women but varied substantially according to race and age, with the strongest associations observed among White and older adults. Among Black adults and those younger than 35 years, BMI was not associated with health care expenditures.

Our findings are consistent with previous studies showing a strong association between BMI and higher health care costs.4–8 Our estimates focused on overweight- and obesity-related health care expenditures at the individual level, after adjustment for different demographic factors and health payer mix. However, in reality, demographic and health payer characteristics frequently differ by body weight such that average expenditures for the same health services may be lower for overweight or obese individuals than for normal-weight individuals; hence, our results are not directly translatable to national health care expenditure estimates attributable to obesity.

Recently, Finkelstein and colleagues8 used data from the 1998 MEPS to estimate national costs attributable to overweight and obesity. Framed from a health payers’ perspective, their study examined the proportions of expenditures attributable to overweight and obesity for each major health payer. In addition, to estimate expenditures for the entire US population, they multiplied the estimates from MEPS by health care expenditures reported by the National Health Accounts to capture national health care costs.8 These investigators estimated the cost attributable to overweight and obesity for the entire US population to be $92.6 billion (in 2002 dollars), or 9% of total costs.

In our study, the rise in health care expenditures associated with higher BMI occurred across all of the major categories of care examined. While hospitalization costs were the largest single source of higher expenditures associated with obesity, obesity led to the largest relative rise in expenditures for prescription medications. These findings suggest that obesity may be accelerating the rise in drug costs in the United States.

Our finding that health care expenditures associated with obesity became progressively higher as adults increased in age was somewhat unexpected. The results of our study are contrary to those of Quesenberry and colleagues,5 who found a weaker association between higher BMIs and health care costs among older patients enrolled in the Kaiser Permanente Medical Care Program. One possible reason for this inconsistency is that the Quesenberry et al. sample may have overrepresented healthy older adults, resulting in underestimations of costs for older obese patients. Nevertheless, previous epidemiological studies have suggested a smaller association between higher BMI and adverse health outcomes among adults of older ages.12,22 This weaker correlation could be attributed to poor nutrition and weight loss as markers of illness in the elderly.

In addition, BMI is a less reliable surrogate for adiposity among older than younger individuals.23 Finally, others have speculated that obese individuals who survive to old age represent a select healthy subgroup and are less likely to suffer from obesity-related complications.24 Our study suggests that while higher BMI may not lead, on average, to higher mortality rates among older adults, it may confer substantial adverse health effects at least in a subgroup of older adults that drive increases in health care costs.

The association between BMI and health care expenditures appears to vary among racial/ethnic groups. The relationship between expenditures and higher BMI was strongest for White adults. Among Hispanic and, particularly, Black adults, BMI was not related to health care costs. In the case of Black adults, this finding appeared to be driven largely by the lack of association between higher BMIs and inpatient expenditures. Reasons for these discrepancies according to race are unclear. One possibility is that we may have missed important associations owing to the modest number of Black adults included in our study.

Conversely, obesity may have different biological effects among Whites and Blacks. Several previous studies have shown a weaker association between higher BMIs and adverse health outcomes and mortality in Black relative to White Americans.10,25,26 Discrepancies in obesity-related health care costs according to race may also be a result of the disparities in care documented in Black and other minority populations.27 In terms of obesity-related conditions, Black adults may not receive care comparable to that received by White adults.

Finally, competing health risks and causes of death such as smoking may be more prevalent among thinner Black adults, blunting the association between higher BMIs and higher health care costs.28 Although our results cannot provide an explanation for the mechanisms underlying the racial differences observed, they highlight the importance of understanding the effects of obesity on different populations.

Given that the rise in obesity disproportionately affects minority populations,3,4 our findings suggest that projections of future health care costs attributable to obesity and cost-effectiveness studies of obesity interventions will need to consider the demographic makeup of the overweight and obese population. Over the next two and a half decades, the number of individuals older than 65 years is expected to rise by 50%.29 Our findings suggest that, as the population ages and life expectancy lengthens, obesity-related costs will probably rise precipitously.

Several limitations of our study should be noted. First, our estimates were based on expenditure data rather than actual health care costs; the actual costs attributable to overweight and obesity may be lower than we observed. Second, while MEPS was designed as a nationally representative survey, the generalizability of our findings is limited by our exclusion of adults among whom height and weight were not ascertained; this group was younger, more likely to be female and White, and less likely to incur health care expenditures. This bias was potentially balanced by other factors that might have caused us to underestimate the costs associated with higher BMI. For example, although the MEPS surveyors verified expenditure data, the survey ultimately relied on respondents to recall or document health service use. Moreover, because height and weight were self-reported, BMI levels were probably underestimated.30 Finally, the survey was conducted in 1998, and patterns of care may have changed since that time.

In summary, our study suggests that mean per capita annual health care costs associated with overweight and obesity are substantial. Expenditures associated with higher BMIs appear to vary according to race/ethnicity and age and to be rising disproportionately among White and older adults. Future studies are needed to examine the factors that drive demographic differences in obesity-related health care costs.

TABLE 1— Unadjusted Median per Capita Health Care Expenditures (December 2003 $) and Interquartile Ranges Across Weight Categories
TABLE 1— Unadjusted Median per Capita Health Care Expenditures (December 2003 $) and Interquartile Ranges Across Weight Categories
 BMI Group, kg/m2
 < 18.518.5–24.925.0–29.930.0–34.935.0–39.9≥ 40
Sample size269451938221538476236
    Total883 (113, 3446)728 (175, 2402)744 (155, 2718)1092 (247, 4028)1072 (222, 4550)1422 (481, 4438)
    Inpatient0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)
    Outpatient hospital baseda0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 7)
    Prescriptions108 (0, 544)64 (0, 400)76 (0, 518)167 (0, 883)245 (14, 981)335 (40, 1137)
    Office visits176 (0, 555)164 (0, 561)156 (0, 545)231 (0, 792)247 (31, 718)315 (72, 802)
    Emergency room visits0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)0 (0, 0)

Note. Values in parentheses are interquartile ranges (25%, 75% respectively).

a Excludes office and emergency room visits.

Support for this study was provided by the National Institute of Diabetes, Digestive, and Kidney Diseases (grant K23DK02962); the Paul Beeson Physician Scholar in Aging Research Program; and the National Institutes of Health.

We thank the National Center for Health Statistics and the Agency for Healthcare Research and Quality for providing the initial data. We also thank Ashley Bourland for assistance with preparation of the article and the figures.

Note. The analyses, interpretations, and conclusions presented in this article are those of the authors and do not reflect those of the National Center for Health Statistics or the Agency for Healthcare Research and Quality.

Human Participant Protection No protocol approval was needed for this study.


1. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA. 1999;282:1523–1529. Crossref, MedlineGoogle Scholar
2. Allison DB, Fontaine KR, Manson JE, Stevens J, VanItallie TB. JAMA. 1999;282:1530–1538. Crossref, MedlineGoogle Scholar
3. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999–2000. JAMA. 2002;288:1723–1727. Crossref, MedlineGoogle Scholar
4. Wolf AM, Colditz GA. Current estimates of the economic cost of obesity in the United States. Obes Res. 1998;6:97–106. Crossref, MedlineGoogle Scholar
5. Quesenberry CPJr, Caan B, Jacobson A. Obesity, health services use, and health care costs among members of a health maintenance organization. Arch Intern Med. 1998;158:466–472. Crossref, MedlineGoogle Scholar
6. Thompson D, Wolf AM. The medical-care cost burden of obesity. Obes Rev. 2001;2:189–197. Crossref, MedlineGoogle Scholar
7. Thompson D, Brown JB, Nichols GA, Elmer PJ, Oster G. Body mass index and future healthcare costs: a retrospective cohort study. Obes Res. 2001;9:210–218. Crossref, MedlineGoogle Scholar
8. Finkelstein EA, Fiebelkorn IC, Wang G. National medical spending attributable to overweight and obesity: how much, and who’s paying? Health Aff [serial online]. 2003;22:219–226. Available at: http://www.healthaffairs.org/WebExclusives/2204Finkelstein.pdf. Accessed October 11, 2004. Google Scholar
9. Solomon CG, Manson JE. Obesity and mortality: a review of the epidemiologic data. Am J Clin Nutr. 1997; 66(suppl 4):1044S–1050S. Crossref, MedlineGoogle Scholar
10. Stevens J. Obesity and mortality in African-Americans. Nutr Rev. 2000;58:346–353. Crossref, MedlineGoogle Scholar
11. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA. 2003;289:187–193. Crossref, MedlineGoogle Scholar
12. Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL. The effect of age on the association between body-mass index and mortality. N Engl J Med. 1998; 338:1–7. Crossref, MedlineGoogle Scholar
13. Heithoff KA, Cuffel BJ, Kennedy S, Peters J. The association between body mass and health care expenditures. Clin Ther. 1997;19:811–820. Crossref, MedlineGoogle Scholar
14. Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The spread of the obesity epidemic in the United States, 1991–1998. JAMA. 1999; 282:1519–1522. Crossref, MedlineGoogle Scholar
15. Medical Expenditure Panel Survey. Available at: http://www.meps.ahrq.gov/. Accessed October 12, 2004. Google Scholar
16. Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. Arch Intern Med. 1998; 159:1855–1867. Google Scholar
17. Duan N, Manning WG, Morris CN, et al. A Comparison of Alternative Models for the Demand for Medical Care. Santa Monica, Calif: RAND Corp; 1982. Google Scholar
18. Goldman DP, Hosek SD, Dixon LS, Sloss EM. The effects of benefit design and managed care on health care costs. J Health Econ. 1995;14:401–418. Crossref, MedlineGoogle Scholar
19. US Dept of Labor, Bureau of Labor Statistics. Consumer Price Index home page. Available at: http://www.bls.gov/cpi/home.htm. Accessed October 12, 2004. Google Scholar
20. Shah BV, Barnwell BG, Hunt PN, La Vange LM. SUDAAN User’s Manual Release With Addendum for SUDAAN Changes From 5.50 to 6.30. Research Triangle Park, NC: Research Triangle Institute; 1992. Google Scholar
21. Koch CG, Freeman DH, Freeman JL. Strategies in the multivariate analysis of data from complex surveys. Int Stat Rev. 1975;43:59–78. CrossrefGoogle Scholar
22. Bender R, Jöckel KH, Trautner C, Spraul M, Berger M. Effect of age on excess mortality in obesity. JAMA. 1999;281:1498–1504. Crossref, MedlineGoogle Scholar
23. Lesser GT, Pierson RN, Pierson RN Jr. Age, body-mass index, and mortality. N Engl J Med. 1998;338:1158–1159. Crossref, MedlineGoogle Scholar
24. Redberg RF. Age, body-mass index, and mortality. N Engl J Med. 1998;338:1159. MedlineGoogle Scholar
25. Folsom AR, Stevens J, Schreiner PJ, McGovern PG. Body mass index, waist/hip ratio, and coronary heart disease incidence in African Americans and whites. Am J Epidemiol. 1998;148:1187–1194. Crossref, MedlineGoogle Scholar
26. Okosun IS, Chandra KM, Choi S, Christman J, Dever GE, Prewitt TE. Hypertension and type 2 diabetes comorbidity in adults in the United States: risk of overall and regional adiposity. Obes Res. 2001;9:1–9. Crossref, MedlineGoogle Scholar
27. Black-white disparities in health care. JAMA. 1990;263:2344–2346. Crossref, MedlineGoogle Scholar
28. Tobacco Use Among U.S. Racial/Ethnic Minority Groups—African Americans, American Indians and Alaska Natives, Asian Americans and Pacific Islanders, and Hispanics: A Report of the Surgeon General. Atlanta, Ga: Centers for Disease Control and Prevention; 1998. Google Scholar
29. US Bureau of the Census. Projected resident population of the United States as of July 1, 2025. Available at: http://www.census.gov/population/www/projections/np_p3.gif. Accessed October 12, 2004. Google Scholar
30. Rowland ML. Self-reported weight and height. Am J Clin Nutr. 1990;52:1125–1133. Crossref, MedlineGoogle Scholar


No related items




Christina C. Wee, MD, MPH, Russell S. Phillips, MD, Anna T. R. Legedza, ScD, Roger B. Davis, ScD, Jane R. Soukup, MS, Graham A. Colditz, MD, DrPH, and Mary Beth Hamel, MD, MPHChristina C. Wee, Russell S. Phillips, Anna T. R. Legedza, Roger B. Davis, Jane R. Soukup, and MaryBeth Hamel are with the Division of General Medicine and Primary Care, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Mass. Graham A. Colditz is with the Channing Laboratory, Brigham and Women’s Hospital and Harvard Medical School. “Health Care Expenditures Associated With Overweight and Obesity Among US Adults: Importance of Age and Race”, American Journal of Public Health 95, no. 1 (January 1, 2005): pp. 159-165.


PMID: 15623877