In the United States, the prevalence of obesity, defined as body mass index (BMI) of at least 30 kg/m2, has increased gradually from 13% of adults in 1960–1962 to 36% in 2009–20101 and is projected to increase to 42% by 2030.2 In addition, 33% of adults have a BMI between 25 and 29.9 kg/m2 and are considered overweight.1 Obesity and overweight are associated with increased risk for many comorbid conditions, including type 2 diabetes mellitus (T2DM), hypertension, dyslipidemia, coronary heart disease, stroke, gallbladder disease, osteoarthritis, sleep apnea, respiratory problems, and endometrial, breast, prostate, and colon cancers.3–6 In addition, higher rates of emotional exhaustion and psychological complaints,7 lower quality of life,8 and mortality9 are correlated with obesity. These can lead to an increased prevalence of associated comorbid conditions, which can cause a significant burden on health care resources and costs.
Medical costs related to obesity in the United States have grown from $52 billion in 199510 to approximately $147 billion in 2008.11 Over this period, the percentage of overall medical costs that are attributable to obesity has risen from 5% to nearly 10%.10–13 Much of this cost growth attributable to obesity stems from associated comorbid conditions. For example, 27% of the increase in overall medical spending in the United States from 1987 to 2001, including 38% of the increase in diabetes costs and 41% of the increase in heart disease costs over that period, was due to obesity.14
A number of prior studies have found higher direct medical and drug costs in overweight and obese individuals than in persons with normal weight.15–20 Additional studies have compared the medical and drug costs of subsets of persons with obesity, defined by BMI levels such as21–25 30.0 to 34.9, 35.0 to 39.9, and more than 40 kg/m2, or found the cost increase per unit increase of BMI.18,20,26 These studies found that direct health care costs and comorbidities increase significantly as BMI increases, particularly more than 27 kg/m2.
Obesity is a significant burden to employers, not only from health care costs but also productivity loss. Two 1994 studies estimated that US business expenditures for sick leave and disability insurance attributable to obesity were $2.4 billion and $800 million, respectively, annually,13 and that obesity was associated with 39.2 million lost work days and contributed $3.9 billion to the cost of lost productivity in the United States.10 A study completed in 2003 placed the cost of lost productive time attributable to obesity at $11.7 billion, but the lost productive time of overweight employees in the study was similar to that of normal-weight employees.27
Studies have found that obese employees are 25% to more than 100% more likely to be absent from work19,28–35 compared with normal-weight employees. In addition, obese employees miss more work and have higher costs in the areas of disability, workers' compensation, sick leave, and other health-related absences.18,23–25,36–41 Previous studies of sick leave costs used salary averages to calculate costs rather than the individual employee's actual salary or actual sick leave payments, and all but two sick leave absence studies19,32 used imputed or self-reported sick leave days.
Comparisons of employee termination and turnover rates between employees with and without obesity were not found, but several studies found that persons with obesity were less likely to be employed,33,42 more likely to take an early retirement,30 more likely to be disqualified from being hired, and less often nominated for supervisor positions.43 Also, studies have reported some employees with obesity having significantly more self-reported work impairment or limitations,5,42,44,45 as well as increased difficulty getting along with coworkers46 and increased presenteeism.8,25,27,34,38,41
Many of the prior studies also found overweight individuals to be a significant burden to employers15,23–25,28,35–38; however, the definition of overweight used is not as consistent in the literature as that of obesity (BMI ≥ 30 kg/m2). Most studies that examine overweight persons use BMI values of 25.0 to 29.9 kg/m2 to define overweight, while several used a lower bound of approximately 27 kg/m2 for overweight, particularly when studying patients with comorbid conditions.18,29,32,47–50 Possibly because BMI is not a perfect measure, there have been inconsistent findings, with cost outcomes being the same or lower in overweight persons compared with that in normal-weight persons. In fact, among all levels of BMI, Burton et al18 found health care costs to be lowest for BMI values between 25 and 27 kg/m2. Nevertheless, BMI values more than 27 kg/m2 have been shown to be associated with increased health risks, including high levels of blood pressure, glucose, and cholesterol.47,50
This study examines three cohorts of employees defined by BMI (kg/m2): BMI < 27, 27 ≤ BMI < 30, and BMI ≥ 30. The 27 ≤ BMI < 30 cohort is further divided into subcohorts on the basis of the presence or absence of comorbidities such as T2DM, hypertension, and dyslipidemia, as described hereafter. The objective of this study was to evaluate the economic burden of overweight and obesity in a large population of employees by comparing all-cause medical and all-cause prescription pharmacy costs; medical costs by place of service; nonimputed sick leave (rare in the literature), short-term disability (STD), long-term disability (LTD), and workers' compensation costs and absence days; turnover rates; and self-reported productivity loss between BMI cohorts and between comorbidity subcohorts. This seems to be the first study to examine such a wide variety of employee-related outcomes together in one study, the first to measure the effects of obesity on employee turnover, and the first to compare employees with BMI between 27 and 30 kg/m2 with those with BMI ≥ 30 kg/m2.
The analytic database used to support this research was extracted from the Human Capital Management Services Research Reference Database (RRDb). The RRDb contains de-identified, integrated information on more than 1,700,000 employees and their dependents from various large employee work groups throughout the United States. The RRDb represents various industries, including communications, transportation, finance, health care, and retail. The database contains 2001–2012 employee-specific information on demographics, salary and payroll, company type, job type, employment status, health insurance claims, and work absence.
Work absence data metrics come from four types of employee absence data: sick leave, STD, LTD, and workers' compensation. These data are collected by HCMS Group from each employer's payroll system, disability insurance carrier, or workers' compensation claims administrator. They provide actual days absent from work, not imputed from medical services, or self-reported.
The study population included employees from the RRDb who had a valid BMI measurement. Body mass index measurements were defined using health risk appraisal data from various employers in the RRDb. The first BMI measurement on record for that employee was used to define the three main study cohorts: “BMI < 27,” “27 ≤ BMI < 30,” and “BMI ≥ 30.” The date of this BMI measurement was used as the employee's index date for the study.
To assess the impact of comorbidities in overweight employees, individuals in the 27 ≤ BMI < 30 cohort were further separated into three subcohorts, using primary, secondary, or tertiary International Classification of Diseases, Ninth Revision, codes for T2DM, hypertension, or dyslipidemia in the medical claims data. Employees with a diagnosis of (or International Classification of Diseases, Ninth Revision, code for) T2DM in the year after the index date were placed into the diabetes with or without hypertension/dyslipidemia subcohort (“T2DM subcohort”). Employees with a diagnosis of hypertension or dyslipidemia (HTN/DL) in the year after the index date and no T2DM diagnosis were placed in the no diabetes with hypertension/dyslipidemia subcohort (“HTN/DL subcohort”). All other employees were placed in the no diabetes/hypertension/dyslipidemia subcohort (“No T2DM/HTN/DL subcohort”).
All employees were required to have a valid BMI measurement. Employees were required to have 12 months of health plan enrollment after the index date and were required to be aged 18 years or older on the index date. Employees were excluded if they had a pregnancy-related medical claim at any point in the medical claims data.
Descriptive Characteristics and Study Outcomes
Descriptive employee characteristics and outcomes were compared between BMI cohorts and comorbidity subcohorts. Descriptive characteristics included age, sex, race, annual salary, tenure (years with employer), exempt status (not eligible for overtime pay), full-time versus part-time, region (as defined by the first digit of the employee's zip code), and Charlson Comorbidity Index.51
Direct health care costs over the 12-month postindex measurement period included employer- and member-paid amounts from all-cause medical and pharmacy insurance claims. In portions of the analysis, medical costs were also divided into seven categories according to the place where the service was provided (inpatient hospital, outpatient hospital or clinic, doctor's office, emergency department, laboratory, other, and unknown).
Work absence costs and days were also summed over the 12-month postindex period for employees who were enrolled in absence plans. Four types of work absence data were examined. Sick leave payments made to employees and days absent were obtained from employer payroll data. Short- and long-term disability payments and days absent came from disability insurance carrier data collected on behalf of the employers. Workers' compensation costs and days absent for work-related injuries and illnesses were obtained from workers' compensation claims data, and workers' compensation costs included medical and pharmacy claims for work-related injuries and payments made to the employee for salary replacement while absent.
Information on productivity at work for a subset of employees was derived from the World Health Organization Health and Work Performance Questionnaire's (HPQ's)52 overall performance scores (self-reported assessments of personal productivity over the past 28 days). The first HPQ completed during the year after the index date was used. Scores range from 0 to 10, with 10 being the highest productivity level. Turnover rates were calculated as the percentage of employees in the study who terminated employment with their current employer between 12 and 18 months after the index date.
Descriptive characteristics were compared with adjustments for multiple comparisons (more than 2 cohorts), using the Marascuilo procedure53 for discrete variables and the Tukey–Kramer procedure54,55 for continuous variables. The Marascuilo and Tukey–Kramer procedures are extensions of chi-squared tests and t tests, respectively. These procedures adjust for multiple comparisons when testing the differences between more than one pair of proportions or means (ie, comparing more than two cohorts).
Regression modeling was used to compare the cost, absence, productivity, and turnover outcomes between the three main BMI cohorts and between the three comorbidity subcohorts of the 27 ≤ BMI < 30 cohort, while adjusting for differences in age, sex, marital status, race, salary, zip code regions, exempt status (similar to salaried vs hourly), and year of the index date. Interaction terms between sex and cohort or subcohort were used to produce separate results for women and men.
Two-part regression models (logistic paired with generalized linear models with gamma distributions and log link functions) were used to model costs and absence days. In the first part of the process, logistic regression was used to model the likelihood of having greater than zero costs (or absences). Second, generalized linear models were used to estimate costs (or absence days) for the portion of the population with greater than zero costs (or absences). The regression coefficients from these models and overall means of the independent variables in the models were combined to produce estimates of the probability of having some cost (or absence) and average costs (or absences) conditional on having some cost (or absence). The probabilities and conditional means were then multiplied together to obtain mean cost and absence estimates for all employees in each cohort.
For other outcomes, two-part modeling was not required. Gamma generalized linear regression with log link functions was used to model productivity scores, and logistic regression was used to model turnover rates. Independent variables were removed from the regression models when necessary to obtain model convergence and fit after examining high pair-wise correlations between variables and by using stepwise variable selection methods.
Each absence cost or days analysis was run on the subset of employees who were eligible for the given type of absence (sick leave, STD, LTD, or workers' compensation). June 2012 Consumer Price Index values for medical services, prescription drugs, or all goods and services were used to adjust costs in the study for inflation.56 Differences in outcomes were considered statistically significant when two-sided P values were less than 0.05. All statistical analyses were conducted using SAS software (version 9.2; SAS Institute, Cary, NC).
Figure 1 depicts the employee selection sequence based on the study inclusion and exclusion criteria and shows that 39,969 employees were in the BMI < 27 cohort, 14,281 were in the 27 ≤ BMI < 30 cohort, and 18,801 were in the BMI ≥ 30 cohort. The 27 ≤ BMI < 30 cohort was then broken into three subcohorts, yielding 9873 employees with no T2DM/HTN/DL, 3859 employees without T2DM but with HTN/DL, and 549 employees with T2DM with or without HTN/DL.
Comparisons of descriptive characteristics for the three main study cohorts are shown in Table 1. Employees in the BMI < 27 cohort were significantly younger (38.8 years) than those in the 27 ≤ BMI < 30 and BMI ≥ 30 cohorts (40.9 years). There was a lower proportion of female employees in the 27 ≤ BMI < 30 cohort (22.9%) than in the BMI < 27 cohort (32.8%) and the BMI ≥ 30 cohort (38.9%). Annual salary was significantly lowest when BMI was highest, averaging $87,604, $83,178, and $65,843 in the BMI < 27, 27 ≤ BMI < 30, BMI ≥ 30 cohorts, respectively. Employees with high BMI were also less likely to be exempt and had higher Charlson Comorbidity Index values.
Demographic information was also available for the three comorbidity subcohorts within the 27 ≤ BMI < 30 cohort (Table 1). Average age and Charlson Comorbidity Index values were the highest in the T2DM subcohort (47.7 years and 1.630, respectively), followed by the HTN/DL subcohort (44.9 years and 0.249, respectively) and the No T2DM/HTN/DL subcohort (38.9 years and 0.089, respectively). Annual salary was highest in the HTN/DL subcohort, and tenure was lowest in the No T2DM/HTN/DL subcohort.
Outputs from this study's regression modeling are included as Supplemental Digital Content 1 (http://links.lww.com/JOM/A148). Adjusted outcomes comparisons between study cohorts (Table 2) revealed that medical, drug, sick leave, STD, and workers' compensation costs and sick leave, STD, and workers' compensation absence days all were higher in cohorts with higher BMI, and all such comparisons between cohorts were significant, except for the STD cost and days comparisons between the BMI < 27 and 27 ≤ BMI < 30 cohorts. Total costs and total days absent in the BMI < 27, 27 ≤ BMI < 30, and BMI ≥ 30 cohorts were $4258, $4873, and $6313 and 4.46, 5.57, and 7.54 days, respectively. The HPQ productivity scores were significantly lowest in BMI ≥ 30 cohort (8.23), but the difference in productivity score between the BMI < 27 (8.34) and 27 ≤ BMI < 30 (8.33) cohorts was not statistically significant. Semiannual turnover rates in each cohort were nearly 4%, but no significant differences were detected between the three cohorts.
All outcomes were also stratified by sex. As shown in Table 2, total costs and total days absent in the BMI < 27, 27 ≤ BMI < 30, and BMI ≥ 30 cohorts for women were $5302, $5946, and $7932 and 5.58, 6.91, and 8.66 days, respectively. For men, costs and total days absent were $3648, $4248, and $4471 and 3.58, 4.61, and 6.95 days, respectively.
Figure 2 provides comparisons between cohorts of adjusted annual medical cost estimates broken out by place of service category for both sexes and for women and men separately. Looking at both sexes combined, the cohorts with higher levels of BMI were associated with significantly higher medical costs in every place of service category. The BMI ≥ 30 cohort had particularly high inpatient costs relative to the other cohorts. When costs by place of service were examined for women and men separately, the BMI ≥ 30 had significantly highest costs in each category, except for unknown place of service costs for women. Note that because of the nonlinearity of the regression models, the sum of the medical costs shown in Figure 2 is slightly different than that shown in the medical cost rows of Table 2.
In the comparisons of both sexes combined (Table 3) between the three subcohorts of the 27 ≤ BMI < 30 cohort, the T2DM subcohort generally had the highest costs and most absence days, and the No T2DM/HTN/DL subcohort generally had the lowest costs and least days absent. This was not the case for LTD, where the T2DM subcohort had no LTD claims during the study period. Between subcohorts, all differences in medical, drug, sick leave, and STD costs and days absent were significant except the differences in STD costs and days between the HTN/DL subcohort and the T2DM subcohort. Total costs and total days absent in the No T2DM/HTN/DL, HTN/DL, and T2DM subcohorts were $3868, $6517, and $9243 and 3.69, 5.41, and 6.36 days, respectively. Productivity scores and turnover rates were similar across all subcohorts.
As shown in Table 3, the subcohort comparisons specifically for women and men revealed that costs and absence days were generally highest in the T2DM subcohort and lowest in the No T2DM/HTN/DL subcohort; however, fewer differences were significant. For women, total costs and total days absent in the No T2DM/HTN/DL, HTN/DL, and T2DM subcohorts were $5498, $7802, and $10,866 and 6.09, 7.06, and 7.95 days, respectively. Total costs and total days absent for men in the No T2DM/HTN/DL, HTN/DL, and T2DM subcohorts were $3398, $5982, and $8817 and 2.79, 4.18, and 5.59 days, respectively.
Figure 3 provides the comparisons of medical costs by place of service for the three subcohorts of the 27 ≤ BMI < 30 cohort. For combined sexes, employees with none of the comorbidities had significantly (P < 0.05) lower medical costs than the other subcohorts in every medical cost place of service category. This was also true for women specifically and was true for men except in other place of service costs. Overall, the T2DM subcohort had significantly higher inpatient, doctor's office, laboratory, and other medical costs than the HTN/DL subcohort. Among women, the T2DM subcohort had significantly higher doctor's office and other place of service costs than the HTN/DL subcohort, and among men, this was true for inpatient hospital, laboratory, and other place of service costs.
As shown in the supplemental digital content (http://links.lww.com/JOM/A148), several of the independent variables used in the regression modeling had significant impacts on the cost, days, turnover, and productivity outcomes. For example, older employees were more costly and had more absence days than younger employees in all categories except sick leave, where they were less costly and had fewer absence days. Older employees also had lower turnover rates and greater self-reported productivity. Most models had one or more race and zip code variables that were significant, but these results varied from one type of outcome to another. Employees with index dates that were later in the study period had lower drug costs and more sick leave, STD, turnover, and self-reported productivity than employees with earlier index dates. Finally exempt employees (usually correlated with higher salaries) had higher LTD costs but fewer sick leave, STD, and workers' compensation absence days and lower turnover and productivity than nonexempt employees.
This study adds to the body of prior research on the employee-related burden of overweight and obesity. Prior studies did not examine as wide a variety of outcomes as have been described herein (direct costs, four kinds of absence costs and days, productivity, and turnover rates). This is the first study to report all-cause turnover rates and seems to be the first to measure all types of health-related absence costs from payroll and claims systems rather than from salary estimates or self-reported information.
Most previous studies have described outcomes on the basis of the World Health Organization's57 BMI groupings: 18.5 to 24.9 (normal weight), 25.0 to 29.9 (overweight), and more than 30 kg/m2 (obesity), sometimes further dividing the obesity group. The finding of Burton et al18 that health care costs are lowest at 25 to 27 kg/m2 shows the need for an examination of outcomes using a cutoff of 27 kg/m2. This is especially true because treatment guidelines for the management of overweight are different for those with BMI of 25 to 26.9 kg/m2, where diet, physical activity, and behavioral modifications are recommended, and 27 to 29.9 kg/m2 where pharmacotherapy may be considered as an adjunct to diet and exercise, in the presence of a comorbidity.58 This study compares outcomes for BMI < 27, 27 ≤ BMI < 30, and the traditional BMI ≥ 30 cohort. Furthermore, this study analyzes the 27 ≤ BMI < 30 cohort more closely, by comparing subcohorts based on important comorbid conditions.
In this study, obesity was associated with significant burden for employers. Nearly all adjusted cost and absence day outcomes were higher among employees with higher BMI values. Medical costs in each place of service category were also significantly higher in each successively higher BMI cohort, with the greatest differences in inpatient medical costs. The HPQ productivity scores were significantly lowest in the BMI ≥ 30 cohort, but employee turnover rates were similar between the three cohorts. In the comparisons between the three subcohorts of the 27 ≤ BMI < 30 cohort, those with T2DM had the highest burden, indicating a subcohort of employees that may warrant special attention. Employees with none of the comorbidities had significantly lower medical costs than the other subcohorts in every place of service category. The T2DM subcohort had significantly higher inpatient, doctor's office, laboratory, and other medical costs than the HTN/DL subcohort. Productivity scores and employee turnover rates were similar across all three subcohorts.
Twenty-six percent of the employees in this study had BMI ≥ 30 kg/m2. This percentage is lower than the 36% found in a representative population of employed and unemployed US adults in prior research.1 It is possible that the reason for the difference is the lower likelihood of being employed among persons with obesity.30,33,42,43 In prior studies of employed populations, the percentage of employees with BMI ≥ 30 kg/m2 varied widely (two populations with 21% to 22%,27,31 six populations with 26% to 29%,19,23,25,31,36,37 and three populations with 35% to 39%.25,38,44
The difference in medical costs between the BMI ≥ 30 cohort and employees in the BMI < 27 cohort (51%) was somewhat higher than the 36% difference between persons in BMI ≥ 30 and 20 ≤ BMI < 25 cohorts shown in prior research.15,16 On the contrary, the difference in drug costs in this study between the BMI ≥ 30 and BMI < 27 cohorts (48%) was smaller than the 77% to 105% differences between the BMI ≥ 30 and 20 ≤ BMI < 25 cohorts shown previously.15,16 This could be due to differences in population characteristics or data sources. Also, cost comparisons between the current study and other prior studies are difficult to make directly because of the tendency of other studies to divide persons with BMI ≥ 30 kg/m2 into groups (eg, 30 to 35, 35 to 40, and more than 40 kg/m2); however, prior studies showed increasing costs for higher levels of BMI as did this study.
Medical cost differences between the BMI ≥ 30 and BMI < 27 cohorts were greatest in the inpatient hospital ($488), outpatient hospital or clinic ($417), and doctor's office ($246) places of service. These results corroborate prior research showing that the difference in direct health care costs between persons with obesity and those with normal weight may be primarily because of more office visits for comorbid conditions and inpatient and outpatient hospital care,12,20,22 though more emergency department visits have also been reported among those with obesity.38
Cawley et al35 (13% to 54%) and Finkelstein et al24 (47%) found that employees with obesity had more self-reported health-related absence days than employees with normal weight, whereas this study found a 69% difference in total days between the BMI ≥ 30 and BMI < 27 cohorts. A review of 11 US studies by Neovius et al40 found that workers with obesity had 1 to 3 additional absence days per person-year than employees with normal weight. This is similar to the 1.43 additional sick leave days and 3.08 additional total absence days the BMI ≥ 30 cohort had when compared with the BMI < 27 cohort in the present study. Prior studies showed incremental self-reported sick leave costs of $200 to $400 among employees with obesity.24,25,38 This range includes the $242 difference between the BMI ≥ 30 and BMI < 27 cohorts found in this study.
Greater differences were found when looking specifically at STD and particularly workers' compensation. This study found a 70% difference in STD days between the BMI ≥ 30 and BMI < 27 cohorts, while two prior studies found employees with obesity had approximately twice as many disability days as employees with normal weight.18,36 The previous study examining workers' compensation costs and days absent found that, depending on level of obesity, employees with obesity had 239% to 700% more workers' compensation days and 390% to 550% more costs than employees with normal weight.37 The present study found a 281% difference in workers' compensation days and a 143% difference in costs between the BMI ≥ 30 and BMI < 27 cohorts, results that are qualitatively similar to the prior study, even though there were differences between studies in the definitions of the low BMI cohort (BMI < 27 kg/m2 vs BMI < 25 kg/m2).
This study did not find significantly higher turnover rates among employees with BMI ≥ 30 kg/m2. This is somewhat dissimilar to the prior finding that employees with BMI ≥ 35 kg/m2 were more likely to take early retirement.30 The current study did, however, find significantly lower self-reported productivity scores in the BMI ≥ 30 cohort, a finding that corroborates earlier results,5,8,25,27,34,38,41,42,44,45 although the magnitude of the difference in scores was small.
Several limitations of this study should be mentioned. First, there may be a tendency in self-report data to underestimate weight and overestimate height. This could lead to BMI values that are biased low and lead to an underestimation of the incremental costs and days associated with higher BMI cohorts.59–62 Next, as in other obesity studies, this study used self-reported productivity information, which can be subject to overreporting and recall bias,63 and used International Classification of Diseases, Ninth Revision, codes from medical insurance claims to identify comorbid conditions and define subcohorts rather than using diagnoses from medical charts. Finally, people with undiagnosed T2DM, dyslipidemia, or hypertension may be in the No T2DM/HTN/DL subcohort, leading to conservative estimates of outcomes differences between subcohorts.
The impact of obesity on an employed population can be costly, not only in medical and pharmacy costs but also in absence from work and reduced productivity at work. Significant costs and absences are also found in overweight employees in the 27 ≤ BMI < 30 cohort who have T2DM, hypertension, or dyslipidemia. Therefore, employers need to take action on weight management if they have not already done so. Because weight loss is a difficult battle, there is a need for access to appropriate weight management strategies and coordination between patients, clinicians, and the workplace for successful outcomes.
1. Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999–2010. JAMA. 2012;307:491–497.
2. Finkelstein EA, Khavjou OA, Thompson H, et al. Obesity and severe obesity forecasts through 2030. Am J Prev Med. 2012;42:563–570.
3. National Institutes of Health. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults—the evidence report. Obes Res. 1998;6(suppl 2):51S–209S.
4. National Task Force on the Prevention and Treatment of Obesity. Overweight, obesity, and health risk. Arch Intern Med. 2000;160:898–904.
5. Hertz RP, Unger AN, McDonald M, et al. The impact of obesity on work limitations and cardiovascular risk factors in the US workforce. J Occup Environ Med. 2004;46:1196–1203.
6. Gregg EW, Cheng YJ, Cadwell BL, et al. Secular trends in cardiovascular disease risk factors according to body mass index in US adults. JAMA. 2005;293:1868–1874.
7. Proper KI, Koppes LLJ, van Zwieten MHJ, Bemelmans WJE. The prevalence of chronic psychological complaints and emotional exhaustion among overweight and obese workers. Int Arch Occup Environ Health. 2012;85:537–545.
8. Cash SW, Beresford SAA, Henderson J, et al. Dietary and physical activity behaviours related to obesity-specific quality of life and work productivity: baseline results from a worksite trial. Br J Nutr. 2012;108:1134–1142.
9. Prospective Studies Collaboration. Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet. 2009;373:1083–1096.
10. Wolf AM, Colditz GA. Current estimates of the economic cost of obesity in the United States. Obes Res. 1998;6:97–106.
11. Finkelstein EA, Trogdon JG, Cohen JW, Dietz W. Annual medical spending attributable to obesity: payer- and service-specific estimates. Health Aff (Millwood). 2009;28:2822–2831.
12. Finkelstein EA, Fiebelkorn IC, Wang G. National medical spending attributable to overweight and obesity: how much, and who's paying? Health Aff (Millwood). 2003;22:w3.219–w3.226.
13. Thompson D, Edelsberg J, Kinsey KL, Oster G. Estimated economic costs of obesity to U.S. business. Am J Health Promot. 1998;13:120–127.
14. Thorpe KE, Florence CS, Howard DH, et al. The impact of obesity on rising medical spending. Health Aff (Millwood). 2004;Suppl Web Exclusives:W4–W6.
15. 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.
16. Sturm R. The effects of obesity, smoking, and drinking on medical problems and costs. Health Aff (Millwood). 2002;21:245–253.
17. Hill RK, Thompson JW, Shaw JL, Pinidiya SD, Card-Higginson P. Self-reported health risks linked to health plan cost and age group. Am J Prev Med. 2009;36:468–474.
18. Burton WN, Chen CY, Schultz AB, et al. The economic costs associated with body mass index in a workplace. J Occup Environ Med. 1998;40:786–792.
19. Bungum T, Satterwhite M, Jackson AW, Morrow JR. The relationship of body mass index, medical costs, and job absenteeism. Am J Health Behav. 2003;27:456–462.
20. Raebel MA, Malone DC, Conner DA, Xu S, Porter JA, Lanty FA. Health services use and health care costs of obese and nonobese individuals. Arch Intern Med. 2004;164:2135–2140.
21. Moriarty JP, Branda ME, Olsen KD, et al. The effects of incremental costs of smoking and obesity on health care costs among adults. J Occup Environ Med. 2012;54:286–291.
22. Wee CC, Phillips RS, Legedza ATR, et al. Health care expenditures associated with overweight and obesity among US adults: importance of age and race. Am J Public Health. 2005;95:159–165.
23. Durden ED, Huse D, Ben-Joseph R, Chu BC. Economic costs of obesity to self-insured employers. J Occup Environ Med. 2008;50:991–997.
24. Finkelstein E, Fiebelkorn IC, Wang G. The costs of obesity among full-time employees. Am J Health Promot. 2005;20:45–51.
25. Finkelstein EA, daCostaDiBonaventura M, Burgess SM, Hale BC. The costs of obesity in the workplace. J Occup Environ Med. 2010;52:971–976.
26. Wang F, McDonald T, Bender J, Reffitt B, Miller A, Edington DW. Association of healthcare costs with per unit body mass index increase. J Occup Environ Med. 2006;48:668–674.
27. Ricci JA, Chee E. Lost productive time associated with excess weight in the US workforce. J Occup Environ Med. 2005;47:1227–1234.
28. Arena VC, Padiyar KR, Burton WN, Schwerha JJ. The impact of body mass index on short-term disability in the workplace. J Occup Environ Med. 2006;48:1118–1124.
29. Soteriades ES, Hauser R, Kawachi I, Christiani DC, Kales SN. Obesity and risk of job disability in male firefighters. Occup Med. 2008;58:245–250.
30. Renna F, Thakur N. Direct and indirect effects of obesity on U.S. labor market outcomes of older working age adults. Soc Sci Med. 2010;71:405–413.
31. Howard JT, Potter LB. An assessment of the relationships between overweight, obesity, related chronic health conditions and worker absenteeism. Obes Res Clin Pract. Available at: http://dx.doi.org/10.1016/j.orcp.2012.09.002
. Published online October 10, 2012. Accessed December 20, 2012.
32. Pai CW, Mullin J, Payne GM, Love J, O'Connell G, Edington DW. Factors associated with incidental sickness absence among employees in one health care system. Am J Health Promot. 2009;24:37–48.
33. Klarenbach S, Padwal R, Chuck A, Jacobs P. Population-based analysis of obesity and workforce participation. Obesity. 2006;14:920–927.
34. Robroek SJW, van den Berg TIJ, Plat JF, Burdorf A. The role of obesity and lifestyle behaviours in a productive workforce. Occup Environ Med. 2011;68:134–139.
35. Cawley J, Rizzo JA, Haas K. Occupation-specific absenteeism costs associated with obesity and morbid obesity. J Occup Environ Med. 2007;49:1317–1324.
36. Tsai SP, Ahmed FS, Wendt JK, Bhojani F, Donnelly RP. The impact of obesity on illness absence and productivity in an industrial population of petrochemical workers. Ann Epidemiol. 2008;18:8–14.
37. Ostbye T, Dement JM, Krause KM. Obesity and workers' compensation: results from the Duke Health and Safety Surveillance System. Arch Intern Med. 2007;167:766–773.
38. Goetzel RZ, Gibson TB, Short ME, et al. A multi-worksite analysis of the relationships among body mass index, medical utilization and worker productivity. J Occup Environ Med. 2010;52(suppl 1):S52–S58.
39. Kowlessar NM, Goetzel RZ, Smith Carls G, Tabrizi MJ, Guindon A. The relationship between 11 health risks and medical and productivity costs for a large employer. J Occup Environ Med. 2011;53:468–477.
40. Neovius K, Johansson K, Kark M, Neovius M. Obesity status and sick leave: a systematic review. Obes Rev. 2009;10:17–27.
41. Striegel RH, Bedrosian R, Wang C. Comparing work productivity in obesity and binge eating. Int J Eat Disord. 2012;45:995–998.
42. Tunceli K, Li K, Williams LK. Long-term effects of obesity on employment and work limitations among U.S. adults, 1986 to 1999. Obesity. 2006;14:1637–1646.
43. Giel KE, Zipfel S, Alizadeh M, et al. Stigmatization of obese individuals by human resource professionals: an experimental study. BMC Public Health. 2012;12:525. doi:10.1186/1471-2458-12-525.
44. Gates DM, Succop P, Brehm BJ, Gillespie GL, Sommers BD. Obesity and presenteeism: the impact of body mass index on workplace productivity. J Occup Environ Med. 2008;50:39–45.
45. Rodbard HW, Fox KM, Grandy S SHIELD Study Group. Impact of obesity on work productivity and role disability in individuals with and at risk for diabetes mellitus. Am J Health Promot. 2009;23:353–360.
46. Pronk NP, Martinson B, Kessler RC, Beck AL, Simon GE, Wang P. The association between work performance and physical activity, cardiorespiratory fitness, and obesity. J Occup Environ Med. 2004;46:19–25.
47. Kannel WB, D'Agostino RB, Cobb JL. Effect of weight on cardiovascular disease. Am J Clin Nutr. 1996;63(suppl):419S–422S.
48. Genco RJ, Grossi SG, Ho A, Nishimura F, Murayama Y. A proposed model linking inflammation to obesity, diabetes, and periodontal infections. J Periodontal. 2005;76:2075–2084.
49. Leiter LA, Abbott D, Campbell NRC, Mendelson R, Ogilvie RI, Chockalingam A. Recommendations on obesity and weight loss. CMAJ. 1999;160(suppl):S7–S12.
50. Rabkin SW, Chen Y, Leiter L, Liu L, Reeder BA Canadian Heart Health Surveys Research Group. Risk factor correlates of body mass index. CMAJ. 1997;157(suppl):S26–S31.
51. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383.
52. Kessler RC, Barber C, Beck A, et al. The World Health Organization Health and Work Performance Questionnaire (HPQ). J Occup Environ Med. 2003;45:156–174.
55. SAS Institute Inc. SAS/STAT® 9.2 User's Guide. The ANOVA Procedure. Cary, NC: SAS Institute Inc; 2008.
57. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO Consultation. Technical Report Series No. 894. Geneva, Switzerland: World Health Organization; 2000. Available at: http://whqlibdoc.who.int/trs/WHO_TRS_894.pdf
. Accessed February 8, 2013.
58. National Institutes of Health, National Heart, Lung, and Blood Institute. The Practical Guide: Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. Bethesda, MD: National Institutes of Health; 2000.
59. Cawley J, Burkhauser RV. Beyond BMI: The Value of More Accurate Measures of Fatness and Obesity in Social Science Research. NBER Working Paper Series 12291. Cambridge, MA: National Bureau of Economic Research; 2006.
60. Rowland ML. Self-reported weight and height. Am J Clin Nutr. 1990;52:1125–1133.
61. Rowland ML. Reporting bias in height and weight data. Stat Bull Metrop Insur Co. 1989;70:2–11.
62. Palta M, Prineas RJ, Berman R, Hannan P. Comparison of self-reported and measured height and weight. Am J Epidemiol. 1982;115:223–230.
63. Prasad M, Wahlqvist P, Shikiar R, Shih YT. A review of self-report instruments measuring health-related work productivity: a patient-reported outcomes perspective. Pharmacoeconomics. 2004;22:225–244.