Research
Population-Based Estimates of Decreases in Quality-Adjusted Life Expectancy Associated with Unhealthy Body Mass Index Haomiao Jia, PhDa Matthew M. Zack, MD, MPHb William W. Thompson, PhDc
ABSTRACT Objective. Being classified as outside the normal range for body mass index (BMI) has been associated with increased risk for chronic health conditions, poor health-related quality of life (HRQOL), and premature death. To assess the impact of BMI on HRQOL and mortality, we compared quality-adjusted life expectancy (QALE) by BMI levels. Methods. We obtained HRQOL data from the 1993–2010 Behavioral Risk Factor Surveillance System and life table estimates from the National Center for Health Statistics national mortality files to estimate QALE among U.S. adults by BMI categories: underweight (BMI ,18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and severely obese (BMI $35.0 kg/m2). Results. In 2010 in the United States, the highest estimated QALE for adults at 18 years of age was 54.1 years for individuals classified as normal weight. The two lowest QALE estimates were for those classified as either underweight (48.9 years) or severely obese (48.2 years). For individuals who were overweight or obese, the QALE estimates fell between those classified as either normal weight (54.1 years) or severely obese (48.2 years). The difference in QALE between adults classified as normal weight and those classified as either overweight or obese was significantly higher among women than among men, irrespective of race/ethnicity. Conclusions. Using population-based data, we found significant differences in QALE loss by BMI category. These findings are valuable for setting national and state targets to reduce health risks associated with severe obesity, and could be used for cost-effectiveness evaluations of weight-reduction interventions.
Columbia University, Mailman School of Public Health and School of Nursing, Department of Biostatistics, New York, NY
a
Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Division of Population Health, Atlanta, GA
b
Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities, Division of Birth Defects and Developmental Disabilities, Atlanta, GA
c
Address correspondence to: William W. Thompson, PhD, Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities, 1600 Clifton Rd. NE, MS-E86, Room 4113, Atlanta, GA 30333; tel. 404-226-8428; e-mail .
Public Health Reports / January–February 2016 / Volume 131
177
178 Research
Being classified as overweight or obese has been associated with an increased risk for a wide range of chronic health conditions, including type 2 diabetes mellitus, heart disease, high cholesterol, hypertension, stroke, and some types of cancer,1–3 as well as for poor healthrelated quality of life (HRQOL) and disability.4,5 Obesity has also been associated with an increased risk for premature death.6,7 In one study, 112,000 cardiovascular disease deaths and 14,000 cancer deaths in the United States were attributed to obesity.6 In addition, a report by the Surgeon General suggested that being classified as overweight was associated with an increased risk for death at a younger age for adults aged 30–64 years.1 A single health index that quantifies the total burden of disease for both morbidity and mortality for any risk factor or health behavior, including being overweight or obese, has been shown to be helpful.8,9 Quality-adjusted life years (QALYs) and quality-adjusted life expectancy (QALE) indices use self-reported preference-based HRQOL items, which assign values for health states from 0 (dead) to 1 (perfect health).10,11 Thus, one year of life lived at an HRQOL score of 0.5 would be equal to 0.5 QALYs, the same as only a half year of life lived in perfect health, or an HRQOL score of 1.0.11 QALE at a specific age is defined and calculated as the sum of QALYs from the age to the end of life.10,11 Estimates of losses in QALE associated with specific risk factors have been found to be helpful for assessing the burden of diseases associated with these risk factors and can be used to determine the costeffectiveness of treatments, interventions, and policies that might reduce the impact of these risk factors on health outcomes.8–12 Several recent studies have examined the total burden of obesity by calculating the decline in QALE or QALYs for the target population due to the increase in obesity prevalence rates for the population.13,14 Stewart and colleagues estimated the decreases in QALE for the U.S. adult population associated with the increases in obesity prevalence from 2005 through 2020.13 Jia and Lubetkin estimated the state-level obesity-related QALY losses in one year per U.S. adult population from 1993 to 2008.14 In both studies, the results represented the effects of obesity on QALE or QALYs for the entire U.S. adult population. These studies did not, however, calculate the losses associated with being merely overweight rather than obese, nor did they differentiate the losses associated with being obese from those associated with being severely obese.13–15 It would be especially helpful to estimate incremental effects across different body weights on QALE, considering that body weight represents a continuum, ranging from underweight to severely obese.8,9 Although sev-
eral studies have compared differences in mortality or life expectancy by level of body mass index (BMI), no study has compared differences in the total burden of disease by BMI levels.6,16–19 The primary objective of this study was to compare the difference in QALE by BMI level and estimate QALE loss associated with BMI categories outside the range considered to be healthy among U.S. adults. We calculated QALE by BMI based on the following five categories: underweight, normal weight, overweight, obese, and severely obese. METHODS We analyzed 1993–2010 data from the Behavioral Risk Factor Surveillance System (BRFSS), a state-based system of probability surveys of noninstitutionalized civilian U.S. adults.20,21 Data included respondents’ BMI derived from self-reported estimates of weight and height without shoes, their leisure-time physical activity status (active or inactive) based on whether or not they reported having engaged in any exercise during the preceding 30 days other than as part of their regular job, their self-assessed general health status on a scale ranging from 1 (excellent) to 5 (poor), and the number of days in the preceding 30 days that they reported each of the following: being physically unhealthy, being mentally unhealthy, and having an activity limitation.22,23 We used a previously published algorithm to estimate preference-based HRQOL measurements for the EuroQol 5-Dimension scale (EQ-5D).24,25 The potential bias in estimates of EQ-5D scores using this method has been estimated to be ,1% relative to using the actual EQ-5D questions11 and had acceptable validity based on previous research.11,14,15 To estimate QALE, we constructed life tables using age-specific mortality data from the all-cause U.S. National Center for Health Statistics mortality files.11,15 We estimated death rates, by sex, race/ethnicity, and age group, by dividing the number of deaths by the population size. Because death rates by the five BMI categories were not available, we used hazard ratios depicting the likelihood of dying among people in the underweight, overweight, obese, and severely obese groups relative to those in the normal-weight group and on the basis of corresponding population proportions. We estimated hazard ratios from the 1997–2004 National Health Interview Survey (NHIS) Linked Mortality Files through December 31, 2006, using the Cox proportional hazards model.15,26 We calculated QALE for each of the five BMI-based weight categories by using the estimated EQ-5D scores and life tables.11,15 We defined differences in QALE between the
Public Health Reports / January–February 2016 / Volume 131
Quality-Adjusted Life Expectancy and Obesity 179
comparing the QALE of physically active individuals vs. physically inactive individuals classified as either overweight or obese. Finally, we examined the trends in QALE associated with being classified as overweight or obese in the United States from 1993 to 2010. We performed data analysis using SAS® version 9.3.27
ormal-weight class and the other four BMI classes as n QALE loss.15 QALE loss quantifies the individual-level health lost due to being underweight, overweight, obese, or severely obese. We estimated the total burden of disease for being overweight or obese for the entire population, or the “population QALE loss” as a result of being overweight or obese, as the difference in QALE using the same mortality rate and HRQOL scores of the normal-weight group for the obese and overweight groups and the QALE of the entire population.15 This index quantifies increases in QALE for the target population if all individuals with a BMI $25 kg/m2 lowered their weight to the normal, healthy BMI range.13,15 We calculated QALE by BMI based on the following five categories: underweight (BMI ,18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and severely obese (BMI $35.0 kg/m2). We estimated the QALE loss for individuals categorized in one of the four unhealthy groups (i.e., underweight, overweight, obese, or severely obese) as the difference between QALE for those classified within normal-weight BMI and QALE for each of the unhealthy groups. We estimated these values for the overall population, by sex, and for two racial/ethnic subgroups (non-Hispanic white and non-Hispanic black). We also examined the impact of physical activity on QALE independent of BMI by
RESULTS In 2010, the age-adjusted HRQOL for those who were severely obese was 0.781, which was 0.085 points lower than for those who were normal weight (0.866). Compared with those who were normal weight, the HRQOL was 0.031 points lower for those who were obese, 0.002 points lower for those who were overweight, and 0.060 points lower for those who were underweight. The life expectancy at 18 years of age was 60.6 years for those who were severely obese and 61.3 years for those who were normal weight. This 0.7-year decrease in life expectancy represents the years of life lost to severe obesity. Similarly, the years of life lost for obesity was 0.5 year, for overweight was 0.2 year, and for underweight was 1.0 year (Table). The estimated QALE among U.S. adults at 18 years of age was 54.1 years among those who were normal weight and 48.9, 53.7, 51.6, and 48.2 years among those who were underweight, overweight, obese, and severely obese, respectively. In other words, compared
Table. Health-related quality of life, life expectancy, and quality-adjusted life expectancy at 18 years of age, and reductions in these factors attributable to having a BMI outside the normal range for U.S. adults, 2010a
BMI (kg/m2) Underweight: ,18.5 Normal weight: 18.5–24.9 Overweight: 25.0–29.9 Obese: 30.0–34.9 Severely obese: $35.0
HRQOLc (SE)
HRQOLc loss (SE)
Life expectancy (SE)
Life expectancy loss (SE)
QALE (SE)
QALE loss (SE)
1.7
0.806 (0.005)
0.060 (0.005)
60.4 (0.2)
1.0 (0.1)
48.9 (0.3)
5.2 (0.1)
134,362
34.3
0.866 (0.002)
Ref.
61.3 (0.2)
Ref.
54.1 (0.2)
Ref.
147,842
36.2
0.864 (0.002)
0.002 (0.001)
61.1 (0.2)
0.2 (0.0)
53.7 (0.2)
0.4 (0.1)
72,001
17.4
0.835 (0.002)
0.031 (0.002)
60.9 (0.2)
0.5 (0.1)
51.6 (0.2)
2.5 (0.1)
42,195
10.4
0.781 (0.003)
0.085 (0.002)
60.6 (0.2)
0.7 (0.1)
48.2 (0.2)
5.9 (0.1)
Weighted percentb
6,238
Number
a Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
Estimated weighted proportion in the population, adjusted for sampling weight and post-stratification weight
b
Age-adjusted average HRQOL scores
c
BMI 5 body mass index HRQOL 5 health-related quality of life SE 5 standard error QALE 5 quality-adjusted life expectancy Ref. 5 referent group
Public Health Reports / January–February 2016 / Volume 131
180 Research
with QALE among 18-year-olds of normal weight, the QALE was 5.9 years lower for those who were severely obese, 2.5 years lower for those who were obese, 0.4 year lower for those who were overweight, and 5.2 years lower for those who were underweight (Table). Because only 1.7% of U.S. adults (1.0% of men and 2.3% of women) were underweight, and many had chronic conditions that likely contributed to their being underweight, we did not report results of the underweight respondents in the remainder of this article; however, data are available upon request. The negative effects of excessive body weight on QALE were much stronger among women than among men. Severe obesity was associated with a 7.6year loss in QALE at 18 years of age among women but only a 3.7-year loss in QALE among men, obesity was associated with a 3.7-year loss in QALE among women but only a 1.4-year loss in QALE among men, and overweight was associated with a 1.4-year loss in QALE among women but a 0.4-year increase in QALE
among men. We also observed some differences in QALE loss between non-Hispanic white and nonHispanic black individuals. The relationship between QALE and body weight was in general slightly stronger among non-Hispanic white than among non-Hispanic black men and women. Among non-Hispanic white men and women, the obesity-associated QALE loss was 2.4 years and the severe obesity-associated QALE loss was 5.9 years; among non-Hispanic black adults, the corresponding obesity- and severe obesity-associated QALE losses were 1.6 and 4.9 years, respectively (Figure 1). With respect to being classified as overweight, the associated QALE losses were the same for both nonHispanic white and non-Hispanic black adults, with a 0.3-year loss. This difference in the magnitude of the impact of body weights on QALE for the two racial/ ethnic groups carried over to subgroups by sex. In general, the differences of obesity-associated QALE losses between non-Hispanic white and non-Hispanic
Figure 1. Quality-adjusted life expectancy loss attributable to body mass index outside the normal rangea among U.S. adults, by sex and race/ethnicity, 2010b
U.S. adults by body mass index (BMI) categories: underweight (BMI ,18.5 kilograms per meter squared [kg/m2]), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and severely obese (BMI $35.0 kg/m2).
a
Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
b
Public Health Reports / January–February 2016 / Volume 131
Quality-Adjusted Life Expectancy and Obesity 181
Figure 2. Quality-adjusted life expectancy among U.S. adults, by body mass index category and leisure-time physical activity classification, 2010a
a Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
BMI 5 body mass index Kg/m2 5 kilograms per meter squared
black individuals were much smaller than the differences between men and women. As might be expected, our analysis of the relationship between physical activity and QALE showed that physically active people had much higher QALE than inactive people in the same weight classes. Among those who were either overweight or obese, the QALE at 18 years of age was 55.0 years for those who were active and 45.7 years for those who were inactive, or a 9.3year gain for being physically active. Among those of normal weight, the QALE was 57.4 years for those who were active and 46.9 years for those who were inactive, or a 10.5-year gain for being physically active (Figure 2). Similar relationships between physical activity status and QALE by weight classification were also observed for both men and women and for non-Hispanic white and non-Hispanic black people. Finally, we estimated population QALE loss as a result of being overweight or obese from 1993 to 2010 and compared it with trends in the prevalence for overweight and obese individuals during the same time period. The population QALE loss increased from 0.7 year in 1993 to 1.2 years in 2010, including increases during this period from 1.1 years to 1.9 years among
women and from 0.1 year to 0.4 year among men, as the prevalence for overweight and obese individuals increased from 49% to 64% (Figure 3). DISCUSSION We examined the difference in QALE between those of normal BMI and those in other BMI classifications, or the individual QALE loss for having a BMI outside the normal range. Our findings generally found that QALE among U.S. adults was inversely related to BMI classification from normal to severe obesity, overall, by sex, and for the two racial/ethnic subgroups we examined (except for a slightly higher QALE among overweight men than among normal-weight men). QALE is a single index that quantifies the overall health loss associated with morbidity and mortality, and this study is the first to estimate QALE for the U.S. adult population by BMI categories.13,14 We found several advantages to using QALE to examine health loss associated with excess weight. For example, we can quantify the incremental adverse health effects with the increasing BMI classifications from normal weight to severe obesity, compare o besity/
Public Health Reports / January–February 2016 / Volume 131
182 Research
overweight-associated health losses in different population subgroups, and compare the burdens of diseases associated with obesity or overweight with other risk factors (e.g., smoking).11,15 This analysis provides evidence for clinical decision making and could be used to analyze the burdens of diseases associated with different excessive BMI categories and the cost-effectiveness evaluation of various weight-reduction treatments or interventions.8,11,12 Although women had both a longer life expectancy and a higher QALE than men, their QALE varied substantially more with their BMI classification. The QALE loss to obesity and severe obesity was more than twice as high among women than among men. Moreover, the QALE among overweight men was 0.4 year higher than among normal-weight men, while the QALE among overweight women was 1.4 years lower than among normal-weight women. This difference was because overweight men had nearly the same HRQOL and life expectancy as normal-weight men, whereas
overweight women had a substantially lower HRQOL and life expectancy than normal-weight women. One possible explanation is that people’s perceptions of their weight status may affect their HRQOL score, and overweight men are less likely to perceive themselves as being overweight and/or being in poor health than are overweight women.28 Another possible explanation for QALE being more related to BMI-based weight classification among women than among men is that BMI values do not necessarily reflect the amount of excess body fat a person may have or account for differences in build.29,30 Part of the differences may be due to the failure of the one-size-fits-all BMI classifications to differentiate between a normal weight for men and a normal weight for women. Estimates of the percentage of men who are normal weight have been found to be lower when based on BMI than when based on other measures, such as body fat percentage.29 Among people who were overweight or obese, QALE was approximately 9.3 years higher among
Figure 3. Trends in population quality-adjusted life expectancy loss attributable to being overweight or obese and prevalence for obese and overweight individuals, United States, 1993–2010a
a Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
QALE 5 quality-adjusted life expectancy
Public Health Reports / January–February 2016 / Volume 131
Quality-Adjusted Life Expectancy and Obesity 183
those who were physically active than among those who were inactive; among people who were normal weight, QALE was approximately 10.5 years higher for those who were physically active vs. inactive. This finding demonstrates that physical activity is positively related to QALE independent of weight, and these 9.3- and 10.5-year gains in QALE for being physically active were substantially higher than the 5.9-year loss in QALE to severe obesity. Clinicians could use these numbers to motivate overweight and obese patients to become more physically active. Strengths and limitations Our study had several strengths. First, we analyzed a large representative sample of individuals who responded to the four CDC HRQOL items from the 2010 BRFSS survey to estimate the health preference scores used in this study. Second, we utilized a large sample of individuals who were followed longitudinally from the NHIS (1997 through 2004 cohorts) to obtain reliable death statistics. And third, when we compared the estimated association between HRQOL and BMI from our study with a study using an independent data source,14 we found the estimates to be similar, adding support for the construct validity of these findings. This study was also subject to several limitations. First, we used cross-sectional data to estimate the associations between obesity and QALE, and all of the reported associations could be bidirectional and should not be assumed to be causal. Second, we could not adjust for other factors (e.g., waist circumference, socioeconomic status, or social determinants of health) in these analyses, although they may have independent effects on QALE; as such, they could affect the current estimates of the associations between obesity and QALE.31,32 Third, the associations among physical fitness, obesity, and QALE could be caused by other unmeasured factors. Fourth, the stationary assumption could not be tested,33 and we assumed that the relationships among BMI, HRQOL, and mortality in each age interval remained constant over time.32,33 Fifth, because BMI is not a measure of body fat, using BMI to classify people’s weight status can introduce measurement error.29,30 A previous study suggests that such measurement error may have caused us to underestimate the extent to which excess weight was associated with QALE loss.12 Sixth, because BRFSS data on weight and height are based on self-reports of survey participants, we could not verify the accuracy of these data. One study found that BMI estimates based on self-reported height and weight were likely to be smaller than BMI values based on measured height and weight.34 Lastly, this study relied on the BRFSS’s
unhealthy days questions to estimate preference-based HRQOL scores rather than direct measurements of these scores. Estimates of QALE loss would likely be smaller than the true values.11,20,21,35 However, the underestimation of estimated QALE loss and population QALE loss from these scores has been estimated to be ,2.5% and 7.0% of that using the actual EQ-5D questions.15 CONCLUSION This study quantified the incremental benefits of reducing body weight from severely obese to obese, then to overweight, and finally to normal weight. These findings provide evidence for setting quantitative targets to reduce health risks from being overweight or obese and can be used for the cost-effectiveness evaluation of various weight-reduction treatments or interventions. In addition, the population QALE loss among U.S. adults that was attributable to excess weight increased substantially from 1993 to 2010 along with the combined prevalence of overweight and obesity. Life expectancy has continued to increase in the U.S. population over time despite an increase in the prevalence of both overweight and obesity.17,36 We hope that a better understanding of these trends and of the factors associated with population QALE loss to obesity or overweight could help policy makers to more accurately assess the effectiveness of policies and interventions designed to promote physical activity and reduce the prevalence of obesity-related chronic diseases, such as diabetes mellitus, on their target populations. Financial support for this study was provided in part by a contract with the Centers for Disease Control and Prevention (CDC) (#200-2011-M-41977). The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of CDC.
REFERENCES 1. Department of Health and Human Services (US). The Surgeon General’s call to action to prevent and decrease overweight and obesity. Rockville (MD): Office of the Surgeon General (US); 2001. 2. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA 1999;282:1523-9. 3. Guh DP, Zhang W, Bansback N, Amarsi Z, Birmingham CL, Anis AH. The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health 2009;9:88. 4. Jia H, Lubetkin EI. The impact of obesity on health-related qualityof-life in the general adult US population. J Public Health (Oxf) 2005;27:156-64. 5. Backholer K, Wong E, Freak-Poli R, Walls HL, Peeters A. Increasing body weight and risk of limitations in activities of daily living: a systematic review and meta-analysis. Obes Rev 2012;13:456-68. 6. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific
Public Health Reports / January–February 2016 / Volume 131
184 Research
7.
8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
excess deaths associated with underweight, overweight, and obesity. JAMA 2007;298:2028-37. Majer IM, Nusselder WJ, Mackenbach JP, Kunst AE. Life expectancy and life expectancy with disability of normal weight, overweight, and obese smokers and nonsmokers in Europe. Obesity (Silver Spring) 2011;19:1451-9. Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Costeffectiveness in health and medicine. New York: Oxford University Press; 1996. The Secretary’s Advisory Committee on National Health Promotion and Disease Prevention Objectives for 2020. Phase I report: recommendations for the framework and format of Healthy People 2020. 2008 [cited 2015 Sep 17]. Available from: https://www .healthypeople.gov/sites/default/files/phaseI_0.pdf Rosenberg MA, Fryback DG, Lawrence WF. Computing populationbased estimates of health-adjusted life expectancy. Med Decis Making 1999;19:90-7. Jia H, Zack MM, Thompson WW. State quality-adjusted life expectancy for U.S. adults from 1993 to 2008. Qual Life Res 2011;20: 853-63. Bhattacharya J, Bundorf MK. The incidence of the healthcare costs of obesity. J Health Econ 2009;28:649-58. Stewart ST, Cutler DM, Rosen AB. Forecasting the effects of obesity and smoking on U.S. life expectancy. N Engl J Med 2009;361:2252-60. Jia H, Lubetkin EI. Obesity-related quality-adjusted life years lost in the U.S. from 1993 to 2008. Am J Prev Med 2010;39:220-7. Jia H, Zack MM, Thompson WW, Dube SR. Quality-adjusted life expectancy (QALE) loss due to smoking in the United States. Qual Life Res 2013;22:27-35. Allison DB, Fontaine KR, Manson JE, Stevens J, Vanitallie TB. Annual deaths attributable to obesity in the United States. JAMA 1999;282:1530-8. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA 2010;303:235-41. Olshansky SJ, Passaro DJ, Hershow RC, Layden J, Carnes BA, Brody J, et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med 2005;352:1138-45. Preston SH, Stokes A, Mehta NK, Cao B. Projecting the effect of changes in smoking and obesity on future life expectancy in the United States. Demography 2014;51:27-49. Department of Health and Human Services (US). Behavioral Risk Factor Surveillance System [cited 2015 Sep 17]. Available from: http://www.cdc.gov/brfss Mokdad AH, Stroup DF, Giles WH. Public health surveillance for behavioral risk factors in a changing environment. Recommenda-
22. 23.
24. 25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36.
tions from the Behavioral Risk Factor Surveillance Team. MMWR Recomm Rep 2003;52(RR-9):1-12. Centers for Disease Control and Prevention (US). Measuring healthy days: population assessment of health-related quality of life. Atlanta: CDC; 2000. Centers for Disease Control and Prevention (US), National Center for Chronic Disease Prevention and Health Promotion, Division of Adult and Community Health. Summary of the Health-Related Quality of Life (HRQOL) Surveillance Expert Panel. Atlanta: CDC; 2008. Jia H, Lubetkin EI, Moriarty DG, Zack MM. A comparison of Healthy Days and EuroQol EQ-5D measures in two US adult samples. Appl Res Qual Life 2007;2:209-21. Jia H, Lubetkin EI. Estimating EuroQol EQ-5D scores from Population Healthy Days data. Med Decis Making 2008;28:491-9. Jia H, Zack MM, Moriarty DG, Fryback DG. Predicting the EuroQol Group’s EQ-5D index from CDC’s “Healthy Days” in a US sample. Med Decis Making 2011;31:174-85. SAS Institute, Inc. SAS®: Version 9.3 for Windows. Cary (NC): SAS Institute, Inc.; 2011. Muennig P, Jia H, Lee R, Lubetkin E. I think therefore I am: perceived ideal weight as a determinant of health. Am J Public Health 2008;98:501-6. McAdams MA, Van Dam RM, Hu FB. Comparison of self-reported and measured BMI as correlates of disease markers in US adults. Obesity (Silver Spring) 2007;15:188-96. Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271-92. Jacobs EJ, Newton CC, Wang Y, Patel AV, McCullough ML, Campbell PT, et al. Waist circumference and all-cause mortality in a large US cohort. Arch Intern Med 2010;170:1293-301. Saydah S, Bullard KM, Cheng Y, Ali MK, Gregg EW, Geiss L, et al. Trends in cardiovascular disease risk factors by obesity level in adults in the United States, NHANES 1999–2010. Obesity (Silver Spring) 2014;22:1888-95. Chiang CL. The life table and its applications. Malabar (FL): Robert E. Krieger Publishers; 1984. Burkhauser RV, Cawley J. Beyond BMI: the value of more accurate measures of fatness and obesity in social science research. J Health Econ 2008;27:519-29. Kennedy AP, Shea JL, Sun G. Comparison of the classification of obesity by BMI vs. dual-energy X-ray absorptiometry in the Newfoundland population. Obesity (Silver Spring) 2009;17:2094-9. Arias E. United States life tables, 2010. Natl Vital Stat Rep 2014 Nov 6;63:1-63.
Public Health Reports / January–February 2016 / Volume 131
Population-Based Estimates of Decreases in Quality-Adjusted Life Expectancy Associated with Unhealthy Body Mass Index Haomiao Jia, PhDa Matthew M. Zack, MD, MPHb William W. Thompson, PhDc
ABSTRACT Objective. Being classified as outside the normal range for body mass index (BMI) has been associated with increased risk for chronic health conditions, poor health-related quality of life (HRQOL), and premature death. To assess the impact of BMI on HRQOL and mortality, we compared quality-adjusted life expectancy (QALE) by BMI levels. Methods. We obtained HRQOL data from the 1993–2010 Behavioral Risk Factor Surveillance System and life table estimates from the National Center for Health Statistics national mortality files to estimate QALE among U.S. adults by BMI categories: underweight (BMI ,18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and severely obese (BMI $35.0 kg/m2). Results. In 2010 in the United States, the highest estimated QALE for adults at 18 years of age was 54.1 years for individuals classified as normal weight. The two lowest QALE estimates were for those classified as either underweight (48.9 years) or severely obese (48.2 years). For individuals who were overweight or obese, the QALE estimates fell between those classified as either normal weight (54.1 years) or severely obese (48.2 years). The difference in QALE between adults classified as normal weight and those classified as either overweight or obese was significantly higher among women than among men, irrespective of race/ethnicity. Conclusions. Using population-based data, we found significant differences in QALE loss by BMI category. These findings are valuable for setting national and state targets to reduce health risks associated with severe obesity, and could be used for cost-effectiveness evaluations of weight-reduction interventions.
Columbia University, Mailman School of Public Health and School of Nursing, Department of Biostatistics, New York, NY
a
Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Division of Population Health, Atlanta, GA
b
Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities, Division of Birth Defects and Developmental Disabilities, Atlanta, GA
c
Address correspondence to: William W. Thompson, PhD, Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities, 1600 Clifton Rd. NE, MS-E86, Room 4113, Atlanta, GA 30333; tel. 404-226-8428; e-mail .
Public Health Reports / January–February 2016 / Volume 131
177
178 Research
Being classified as overweight or obese has been associated with an increased risk for a wide range of chronic health conditions, including type 2 diabetes mellitus, heart disease, high cholesterol, hypertension, stroke, and some types of cancer,1–3 as well as for poor healthrelated quality of life (HRQOL) and disability.4,5 Obesity has also been associated with an increased risk for premature death.6,7 In one study, 112,000 cardiovascular disease deaths and 14,000 cancer deaths in the United States were attributed to obesity.6 In addition, a report by the Surgeon General suggested that being classified as overweight was associated with an increased risk for death at a younger age for adults aged 30–64 years.1 A single health index that quantifies the total burden of disease for both morbidity and mortality for any risk factor or health behavior, including being overweight or obese, has been shown to be helpful.8,9 Quality-adjusted life years (QALYs) and quality-adjusted life expectancy (QALE) indices use self-reported preference-based HRQOL items, which assign values for health states from 0 (dead) to 1 (perfect health).10,11 Thus, one year of life lived at an HRQOL score of 0.5 would be equal to 0.5 QALYs, the same as only a half year of life lived in perfect health, or an HRQOL score of 1.0.11 QALE at a specific age is defined and calculated as the sum of QALYs from the age to the end of life.10,11 Estimates of losses in QALE associated with specific risk factors have been found to be helpful for assessing the burden of diseases associated with these risk factors and can be used to determine the costeffectiveness of treatments, interventions, and policies that might reduce the impact of these risk factors on health outcomes.8–12 Several recent studies have examined the total burden of obesity by calculating the decline in QALE or QALYs for the target population due to the increase in obesity prevalence rates for the population.13,14 Stewart and colleagues estimated the decreases in QALE for the U.S. adult population associated with the increases in obesity prevalence from 2005 through 2020.13 Jia and Lubetkin estimated the state-level obesity-related QALY losses in one year per U.S. adult population from 1993 to 2008.14 In both studies, the results represented the effects of obesity on QALE or QALYs for the entire U.S. adult population. These studies did not, however, calculate the losses associated with being merely overweight rather than obese, nor did they differentiate the losses associated with being obese from those associated with being severely obese.13–15 It would be especially helpful to estimate incremental effects across different body weights on QALE, considering that body weight represents a continuum, ranging from underweight to severely obese.8,9 Although sev-
eral studies have compared differences in mortality or life expectancy by level of body mass index (BMI), no study has compared differences in the total burden of disease by BMI levels.6,16–19 The primary objective of this study was to compare the difference in QALE by BMI level and estimate QALE loss associated with BMI categories outside the range considered to be healthy among U.S. adults. We calculated QALE by BMI based on the following five categories: underweight, normal weight, overweight, obese, and severely obese. METHODS We analyzed 1993–2010 data from the Behavioral Risk Factor Surveillance System (BRFSS), a state-based system of probability surveys of noninstitutionalized civilian U.S. adults.20,21 Data included respondents’ BMI derived from self-reported estimates of weight and height without shoes, their leisure-time physical activity status (active or inactive) based on whether or not they reported having engaged in any exercise during the preceding 30 days other than as part of their regular job, their self-assessed general health status on a scale ranging from 1 (excellent) to 5 (poor), and the number of days in the preceding 30 days that they reported each of the following: being physically unhealthy, being mentally unhealthy, and having an activity limitation.22,23 We used a previously published algorithm to estimate preference-based HRQOL measurements for the EuroQol 5-Dimension scale (EQ-5D).24,25 The potential bias in estimates of EQ-5D scores using this method has been estimated to be ,1% relative to using the actual EQ-5D questions11 and had acceptable validity based on previous research.11,14,15 To estimate QALE, we constructed life tables using age-specific mortality data from the all-cause U.S. National Center for Health Statistics mortality files.11,15 We estimated death rates, by sex, race/ethnicity, and age group, by dividing the number of deaths by the population size. Because death rates by the five BMI categories were not available, we used hazard ratios depicting the likelihood of dying among people in the underweight, overweight, obese, and severely obese groups relative to those in the normal-weight group and on the basis of corresponding population proportions. We estimated hazard ratios from the 1997–2004 National Health Interview Survey (NHIS) Linked Mortality Files through December 31, 2006, using the Cox proportional hazards model.15,26 We calculated QALE for each of the five BMI-based weight categories by using the estimated EQ-5D scores and life tables.11,15 We defined differences in QALE between the
Public Health Reports / January–February 2016 / Volume 131
Quality-Adjusted Life Expectancy and Obesity 179
comparing the QALE of physically active individuals vs. physically inactive individuals classified as either overweight or obese. Finally, we examined the trends in QALE associated with being classified as overweight or obese in the United States from 1993 to 2010. We performed data analysis using SAS® version 9.3.27
ormal-weight class and the other four BMI classes as n QALE loss.15 QALE loss quantifies the individual-level health lost due to being underweight, overweight, obese, or severely obese. We estimated the total burden of disease for being overweight or obese for the entire population, or the “population QALE loss” as a result of being overweight or obese, as the difference in QALE using the same mortality rate and HRQOL scores of the normal-weight group for the obese and overweight groups and the QALE of the entire population.15 This index quantifies increases in QALE for the target population if all individuals with a BMI $25 kg/m2 lowered their weight to the normal, healthy BMI range.13,15 We calculated QALE by BMI based on the following five categories: underweight (BMI ,18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and severely obese (BMI $35.0 kg/m2). We estimated the QALE loss for individuals categorized in one of the four unhealthy groups (i.e., underweight, overweight, obese, or severely obese) as the difference between QALE for those classified within normal-weight BMI and QALE for each of the unhealthy groups. We estimated these values for the overall population, by sex, and for two racial/ethnic subgroups (non-Hispanic white and non-Hispanic black). We also examined the impact of physical activity on QALE independent of BMI by
RESULTS In 2010, the age-adjusted HRQOL for those who were severely obese was 0.781, which was 0.085 points lower than for those who were normal weight (0.866). Compared with those who were normal weight, the HRQOL was 0.031 points lower for those who were obese, 0.002 points lower for those who were overweight, and 0.060 points lower for those who were underweight. The life expectancy at 18 years of age was 60.6 years for those who were severely obese and 61.3 years for those who were normal weight. This 0.7-year decrease in life expectancy represents the years of life lost to severe obesity. Similarly, the years of life lost for obesity was 0.5 year, for overweight was 0.2 year, and for underweight was 1.0 year (Table). The estimated QALE among U.S. adults at 18 years of age was 54.1 years among those who were normal weight and 48.9, 53.7, 51.6, and 48.2 years among those who were underweight, overweight, obese, and severely obese, respectively. In other words, compared
Table. Health-related quality of life, life expectancy, and quality-adjusted life expectancy at 18 years of age, and reductions in these factors attributable to having a BMI outside the normal range for U.S. adults, 2010a
BMI (kg/m2) Underweight: ,18.5 Normal weight: 18.5–24.9 Overweight: 25.0–29.9 Obese: 30.0–34.9 Severely obese: $35.0
HRQOLc (SE)
HRQOLc loss (SE)
Life expectancy (SE)
Life expectancy loss (SE)
QALE (SE)
QALE loss (SE)
1.7
0.806 (0.005)
0.060 (0.005)
60.4 (0.2)
1.0 (0.1)
48.9 (0.3)
5.2 (0.1)
134,362
34.3
0.866 (0.002)
Ref.
61.3 (0.2)
Ref.
54.1 (0.2)
Ref.
147,842
36.2
0.864 (0.002)
0.002 (0.001)
61.1 (0.2)
0.2 (0.0)
53.7 (0.2)
0.4 (0.1)
72,001
17.4
0.835 (0.002)
0.031 (0.002)
60.9 (0.2)
0.5 (0.1)
51.6 (0.2)
2.5 (0.1)
42,195
10.4
0.781 (0.003)
0.085 (0.002)
60.6 (0.2)
0.7 (0.1)
48.2 (0.2)
5.9 (0.1)
Weighted percentb
6,238
Number
a Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
Estimated weighted proportion in the population, adjusted for sampling weight and post-stratification weight
b
Age-adjusted average HRQOL scores
c
BMI 5 body mass index HRQOL 5 health-related quality of life SE 5 standard error QALE 5 quality-adjusted life expectancy Ref. 5 referent group
Public Health Reports / January–February 2016 / Volume 131
180 Research
with QALE among 18-year-olds of normal weight, the QALE was 5.9 years lower for those who were severely obese, 2.5 years lower for those who were obese, 0.4 year lower for those who were overweight, and 5.2 years lower for those who were underweight (Table). Because only 1.7% of U.S. adults (1.0% of men and 2.3% of women) were underweight, and many had chronic conditions that likely contributed to their being underweight, we did not report results of the underweight respondents in the remainder of this article; however, data are available upon request. The negative effects of excessive body weight on QALE were much stronger among women than among men. Severe obesity was associated with a 7.6year loss in QALE at 18 years of age among women but only a 3.7-year loss in QALE among men, obesity was associated with a 3.7-year loss in QALE among women but only a 1.4-year loss in QALE among men, and overweight was associated with a 1.4-year loss in QALE among women but a 0.4-year increase in QALE
among men. We also observed some differences in QALE loss between non-Hispanic white and nonHispanic black individuals. The relationship between QALE and body weight was in general slightly stronger among non-Hispanic white than among non-Hispanic black men and women. Among non-Hispanic white men and women, the obesity-associated QALE loss was 2.4 years and the severe obesity-associated QALE loss was 5.9 years; among non-Hispanic black adults, the corresponding obesity- and severe obesity-associated QALE losses were 1.6 and 4.9 years, respectively (Figure 1). With respect to being classified as overweight, the associated QALE losses were the same for both nonHispanic white and non-Hispanic black adults, with a 0.3-year loss. This difference in the magnitude of the impact of body weights on QALE for the two racial/ ethnic groups carried over to subgroups by sex. In general, the differences of obesity-associated QALE losses between non-Hispanic white and non-Hispanic
Figure 1. Quality-adjusted life expectancy loss attributable to body mass index outside the normal rangea among U.S. adults, by sex and race/ethnicity, 2010b
U.S. adults by body mass index (BMI) categories: underweight (BMI ,18.5 kilograms per meter squared [kg/m2]), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and severely obese (BMI $35.0 kg/m2).
a
Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
b
Public Health Reports / January–February 2016 / Volume 131
Quality-Adjusted Life Expectancy and Obesity 181
Figure 2. Quality-adjusted life expectancy among U.S. adults, by body mass index category and leisure-time physical activity classification, 2010a
a Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
BMI 5 body mass index Kg/m2 5 kilograms per meter squared
black individuals were much smaller than the differences between men and women. As might be expected, our analysis of the relationship between physical activity and QALE showed that physically active people had much higher QALE than inactive people in the same weight classes. Among those who were either overweight or obese, the QALE at 18 years of age was 55.0 years for those who were active and 45.7 years for those who were inactive, or a 9.3year gain for being physically active. Among those of normal weight, the QALE was 57.4 years for those who were active and 46.9 years for those who were inactive, or a 10.5-year gain for being physically active (Figure 2). Similar relationships between physical activity status and QALE by weight classification were also observed for both men and women and for non-Hispanic white and non-Hispanic black people. Finally, we estimated population QALE loss as a result of being overweight or obese from 1993 to 2010 and compared it with trends in the prevalence for overweight and obese individuals during the same time period. The population QALE loss increased from 0.7 year in 1993 to 1.2 years in 2010, including increases during this period from 1.1 years to 1.9 years among
women and from 0.1 year to 0.4 year among men, as the prevalence for overweight and obese individuals increased from 49% to 64% (Figure 3). DISCUSSION We examined the difference in QALE between those of normal BMI and those in other BMI classifications, or the individual QALE loss for having a BMI outside the normal range. Our findings generally found that QALE among U.S. adults was inversely related to BMI classification from normal to severe obesity, overall, by sex, and for the two racial/ethnic subgroups we examined (except for a slightly higher QALE among overweight men than among normal-weight men). QALE is a single index that quantifies the overall health loss associated with morbidity and mortality, and this study is the first to estimate QALE for the U.S. adult population by BMI categories.13,14 We found several advantages to using QALE to examine health loss associated with excess weight. For example, we can quantify the incremental adverse health effects with the increasing BMI classifications from normal weight to severe obesity, compare o besity/
Public Health Reports / January–February 2016 / Volume 131
182 Research
overweight-associated health losses in different population subgroups, and compare the burdens of diseases associated with obesity or overweight with other risk factors (e.g., smoking).11,15 This analysis provides evidence for clinical decision making and could be used to analyze the burdens of diseases associated with different excessive BMI categories and the cost-effectiveness evaluation of various weight-reduction treatments or interventions.8,11,12 Although women had both a longer life expectancy and a higher QALE than men, their QALE varied substantially more with their BMI classification. The QALE loss to obesity and severe obesity was more than twice as high among women than among men. Moreover, the QALE among overweight men was 0.4 year higher than among normal-weight men, while the QALE among overweight women was 1.4 years lower than among normal-weight women. This difference was because overweight men had nearly the same HRQOL and life expectancy as normal-weight men, whereas
overweight women had a substantially lower HRQOL and life expectancy than normal-weight women. One possible explanation is that people’s perceptions of their weight status may affect their HRQOL score, and overweight men are less likely to perceive themselves as being overweight and/or being in poor health than are overweight women.28 Another possible explanation for QALE being more related to BMI-based weight classification among women than among men is that BMI values do not necessarily reflect the amount of excess body fat a person may have or account for differences in build.29,30 Part of the differences may be due to the failure of the one-size-fits-all BMI classifications to differentiate between a normal weight for men and a normal weight for women. Estimates of the percentage of men who are normal weight have been found to be lower when based on BMI than when based on other measures, such as body fat percentage.29 Among people who were overweight or obese, QALE was approximately 9.3 years higher among
Figure 3. Trends in population quality-adjusted life expectancy loss attributable to being overweight or obese and prevalence for obese and overweight individuals, United States, 1993–2010a
a Data sources: 1993–2007 Behavioral Risk Factor Surveillance System (http://www.cdc.gov/brfss/annual_data/annual_data.htm), 1997–2001 National Health Interview Survey Linked Mortality Files (http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm).
QALE 5 quality-adjusted life expectancy
Public Health Reports / January–February 2016 / Volume 131
Quality-Adjusted Life Expectancy and Obesity 183
those who were physically active than among those who were inactive; among people who were normal weight, QALE was approximately 10.5 years higher for those who were physically active vs. inactive. This finding demonstrates that physical activity is positively related to QALE independent of weight, and these 9.3- and 10.5-year gains in QALE for being physically active were substantially higher than the 5.9-year loss in QALE to severe obesity. Clinicians could use these numbers to motivate overweight and obese patients to become more physically active. Strengths and limitations Our study had several strengths. First, we analyzed a large representative sample of individuals who responded to the four CDC HRQOL items from the 2010 BRFSS survey to estimate the health preference scores used in this study. Second, we utilized a large sample of individuals who were followed longitudinally from the NHIS (1997 through 2004 cohorts) to obtain reliable death statistics. And third, when we compared the estimated association between HRQOL and BMI from our study with a study using an independent data source,14 we found the estimates to be similar, adding support for the construct validity of these findings. This study was also subject to several limitations. First, we used cross-sectional data to estimate the associations between obesity and QALE, and all of the reported associations could be bidirectional and should not be assumed to be causal. Second, we could not adjust for other factors (e.g., waist circumference, socioeconomic status, or social determinants of health) in these analyses, although they may have independent effects on QALE; as such, they could affect the current estimates of the associations between obesity and QALE.31,32 Third, the associations among physical fitness, obesity, and QALE could be caused by other unmeasured factors. Fourth, the stationary assumption could not be tested,33 and we assumed that the relationships among BMI, HRQOL, and mortality in each age interval remained constant over time.32,33 Fifth, because BMI is not a measure of body fat, using BMI to classify people’s weight status can introduce measurement error.29,30 A previous study suggests that such measurement error may have caused us to underestimate the extent to which excess weight was associated with QALE loss.12 Sixth, because BRFSS data on weight and height are based on self-reports of survey participants, we could not verify the accuracy of these data. One study found that BMI estimates based on self-reported height and weight were likely to be smaller than BMI values based on measured height and weight.34 Lastly, this study relied on the BRFSS’s
unhealthy days questions to estimate preference-based HRQOL scores rather than direct measurements of these scores. Estimates of QALE loss would likely be smaller than the true values.11,20,21,35 However, the underestimation of estimated QALE loss and population QALE loss from these scores has been estimated to be ,2.5% and 7.0% of that using the actual EQ-5D questions.15 CONCLUSION This study quantified the incremental benefits of reducing body weight from severely obese to obese, then to overweight, and finally to normal weight. These findings provide evidence for setting quantitative targets to reduce health risks from being overweight or obese and can be used for the cost-effectiveness evaluation of various weight-reduction treatments or interventions. In addition, the population QALE loss among U.S. adults that was attributable to excess weight increased substantially from 1993 to 2010 along with the combined prevalence of overweight and obesity. Life expectancy has continued to increase in the U.S. population over time despite an increase in the prevalence of both overweight and obesity.17,36 We hope that a better understanding of these trends and of the factors associated with population QALE loss to obesity or overweight could help policy makers to more accurately assess the effectiveness of policies and interventions designed to promote physical activity and reduce the prevalence of obesity-related chronic diseases, such as diabetes mellitus, on their target populations. Financial support for this study was provided in part by a contract with the Centers for Disease Control and Prevention (CDC) (#200-2011-M-41977). The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of CDC.
REFERENCES 1. Department of Health and Human Services (US). The Surgeon General’s call to action to prevent and decrease overweight and obesity. Rockville (MD): Office of the Surgeon General (US); 2001. 2. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA 1999;282:1523-9. 3. Guh DP, Zhang W, Bansback N, Amarsi Z, Birmingham CL, Anis AH. The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health 2009;9:88. 4. Jia H, Lubetkin EI. The impact of obesity on health-related qualityof-life in the general adult US population. J Public Health (Oxf) 2005;27:156-64. 5. Backholer K, Wong E, Freak-Poli R, Walls HL, Peeters A. Increasing body weight and risk of limitations in activities of daily living: a systematic review and meta-analysis. Obes Rev 2012;13:456-68. 6. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific
Public Health Reports / January–February 2016 / Volume 131
184 Research
7.
8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
excess deaths associated with underweight, overweight, and obesity. JAMA 2007;298:2028-37. Majer IM, Nusselder WJ, Mackenbach JP, Kunst AE. Life expectancy and life expectancy with disability of normal weight, overweight, and obese smokers and nonsmokers in Europe. Obesity (Silver Spring) 2011;19:1451-9. Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Costeffectiveness in health and medicine. New York: Oxford University Press; 1996. The Secretary’s Advisory Committee on National Health Promotion and Disease Prevention Objectives for 2020. Phase I report: recommendations for the framework and format of Healthy People 2020. 2008 [cited 2015 Sep 17]. Available from: https://www .healthypeople.gov/sites/default/files/phaseI_0.pdf Rosenberg MA, Fryback DG, Lawrence WF. Computing populationbased estimates of health-adjusted life expectancy. Med Decis Making 1999;19:90-7. Jia H, Zack MM, Thompson WW. State quality-adjusted life expectancy for U.S. adults from 1993 to 2008. Qual Life Res 2011;20: 853-63. Bhattacharya J, Bundorf MK. The incidence of the healthcare costs of obesity. J Health Econ 2009;28:649-58. Stewart ST, Cutler DM, Rosen AB. Forecasting the effects of obesity and smoking on U.S. life expectancy. N Engl J Med 2009;361:2252-60. Jia H, Lubetkin EI. Obesity-related quality-adjusted life years lost in the U.S. from 1993 to 2008. Am J Prev Med 2010;39:220-7. Jia H, Zack MM, Thompson WW, Dube SR. Quality-adjusted life expectancy (QALE) loss due to smoking in the United States. Qual Life Res 2013;22:27-35. Allison DB, Fontaine KR, Manson JE, Stevens J, Vanitallie TB. Annual deaths attributable to obesity in the United States. JAMA 1999;282:1530-8. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA 2010;303:235-41. Olshansky SJ, Passaro DJ, Hershow RC, Layden J, Carnes BA, Brody J, et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med 2005;352:1138-45. Preston SH, Stokes A, Mehta NK, Cao B. Projecting the effect of changes in smoking and obesity on future life expectancy in the United States. Demography 2014;51:27-49. Department of Health and Human Services (US). Behavioral Risk Factor Surveillance System [cited 2015 Sep 17]. Available from: http://www.cdc.gov/brfss Mokdad AH, Stroup DF, Giles WH. Public health surveillance for behavioral risk factors in a changing environment. Recommenda-
22. 23.
24. 25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36.
tions from the Behavioral Risk Factor Surveillance Team. MMWR Recomm Rep 2003;52(RR-9):1-12. Centers for Disease Control and Prevention (US). Measuring healthy days: population assessment of health-related quality of life. Atlanta: CDC; 2000. Centers for Disease Control and Prevention (US), National Center for Chronic Disease Prevention and Health Promotion, Division of Adult and Community Health. Summary of the Health-Related Quality of Life (HRQOL) Surveillance Expert Panel. Atlanta: CDC; 2008. Jia H, Lubetkin EI, Moriarty DG, Zack MM. A comparison of Healthy Days and EuroQol EQ-5D measures in two US adult samples. Appl Res Qual Life 2007;2:209-21. Jia H, Lubetkin EI. Estimating EuroQol EQ-5D scores from Population Healthy Days data. Med Decis Making 2008;28:491-9. Jia H, Zack MM, Moriarty DG, Fryback DG. Predicting the EuroQol Group’s EQ-5D index from CDC’s “Healthy Days” in a US sample. Med Decis Making 2011;31:174-85. SAS Institute, Inc. SAS®: Version 9.3 for Windows. Cary (NC): SAS Institute, Inc.; 2011. Muennig P, Jia H, Lee R, Lubetkin E. I think therefore I am: perceived ideal weight as a determinant of health. Am J Public Health 2008;98:501-6. McAdams MA, Van Dam RM, Hu FB. Comparison of self-reported and measured BMI as correlates of disease markers in US adults. Obesity (Silver Spring) 2007;15:188-96. Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271-92. Jacobs EJ, Newton CC, Wang Y, Patel AV, McCullough ML, Campbell PT, et al. Waist circumference and all-cause mortality in a large US cohort. Arch Intern Med 2010;170:1293-301. Saydah S, Bullard KM, Cheng Y, Ali MK, Gregg EW, Geiss L, et al. Trends in cardiovascular disease risk factors by obesity level in adults in the United States, NHANES 1999–2010. Obesity (Silver Spring) 2014;22:1888-95. Chiang CL. The life table and its applications. Malabar (FL): Robert E. Krieger Publishers; 1984. Burkhauser RV, Cawley J. Beyond BMI: the value of more accurate measures of fatness and obesity in social science research. J Health Econ 2008;27:519-29. Kennedy AP, Shea JL, Sun G. Comparison of the classification of obesity by BMI vs. dual-energy X-ray absorptiometry in the Newfoundland population. Obesity (Silver Spring) 2009;17:2094-9. Arias E. United States life tables, 2010. Natl Vital Stat Rep 2014 Nov 6;63:1-63.
Public Health Reports / January–February 2016 / Volume 131
