The Costs of Obesity Among Full-time Employees

September/October 2005, Vol. 20, No. 1 45

T H E S C I E N C E O F H E A L T H P R O M O T I O N

Financial Analysis; Weight Control

The Costs of Obesity Among Full-time EmployeesEric Finkelstein, PhD, MHA; Ian C. Fiebelkorn, BA; Guijing Wang, PhD

Abstract

Purpose. To quantify annual costs attributable to obesity, including both increasedmedical expenditures and absenteeism, separately for overweight and three categories of obe-sity (i.e., obesity grades I, II, and III) among men and women with full-time employment.

Design. Standard econometric methods were used to separately estimate overweight andobesity-attributable medical expenditures and absenteeism.

Setting. The civilian noninstitutionalized population of the United States.Sample. Two nationally representative and publicly available datasets (with response

rates of at least 60%) were restricted to participants 18 to 64 years old and employed full-time for the entire year. The final datasets used to estimate obesity-attributable medical ex-penditures and absenteeism included 20,329 and 25,427 adults, respectively.

Measures. Annual medical expenditures and missed work days due to illness or injury.Analysis Results. Overweight and obesity-attributable costs range from $175 per year for

overweight male employees to $2485 per year for grade-II obese female employees. The costsof obesity (excluding overweight) at a firm with 1000 employees are estimated to be$285,000 per year.

Conclusions. Obesity results in significant increases in medical expenditures and absen-teeism among full-time employees. Approximately 30% of the total costs result from in-creased absenteeism, and although those with grade-III obesity represent only 3% of theemployed population, they account for 21% of the costs due to obesity. These estimates donot consider other potential costs associated with obesity, including disability and presentee-ism. (Am J Health Promot 2005;20[1]:45–51.)

Key Words: Obesity, Cost, Employer, Prevention Research. Manuscript format:research; Research purpose: modeling/relationship testing; Study design: quasi-experimental; Outcome measure: absenteeism and other financial/economic;Setting: workplace; Health focus: weight control; Strategy: policy; Target popula-tion: adults; Target population circumstances: education/income level, geo-graphic location, and race/ethnicity

Eric Finkelstein, PhD, MHA; and Ian C. Fiebelkorn, BA, are with RTI InternationalHealth, Social and Economics Research, Research Triangle Park, North Carolina. GuijingWang, PhD, is with Centers for Disease Control and Prevention, Atlanta, Georgia.

Send reprint requests to Eric Finkelstein, RTI International Health, Social and Economics Research,Hobbs Building, 3040 Cornwallis Road, Research Triangle Park, NC 27709-2194; [email protected]

This manuscript was submitted May 6, 2005; the manuscript was accepted for publication May 28, 2005.

Copyright q 2005 by American Journal of Health Promotion, Inc.0890-1171/05/$5.00 1 0

INTRODUCTION

The increased prevalence of obesi-ty and obesity-related diseases amongAmericans is well documented.1–3

There is also a growing body of liter-ature on the adverse financial conse-

quences associated with obesity,which largely focuses on the in-creased medical expenditures that re-sult from excess weight. Sturm4 usedregression analysis to show that obeseadults (body mass index [BMI] 301kg/m2) incur annual medical expen-

ditures that are, on average, 36%higher than expenditures for normal-weight adults (BMI 18.5–25 kg/m2).Finkelstein and colleagues5 foundthat overweight and obesity accountfor 9.1% of total annual U.S. medicalexpenditures.

For employers, an increasinglyobese workforce and correspondingincreases in the cost of health carepresent a significant obstacle to pro-viding affordable health insurance toemployees. Moreover, in addition torising medical expenditures, obeseemployees are absent from workmore often than their nonobesecounterparts. Thompson and col-leagues6 compared mean missedwork days for employees with a BMIgreater than 29 kg/m2 with daysmissed for employees with a BMI lessthan 25 kg/m2. Although the studydid not control for other potentialdifferences between obese and nor-mal-weight employees (e.g., income,education), the results suggest thatobese male employees miss up to 2.7more days per year than normal-weight males, and obese female em-ployees miss up to 5.1 more days peryear than normal-weight females. Al-though some of the costs associatedwith obesity may be passed along tothe employees (both obese and non-obese) in the form of lower wagesand/or higher insurance premiums,employers will ultimately bear someof the burden and thus have a finan-cial motivation to search for strate-gies to reduce these costs.

PURPOSE

In the present analysis, we expandon prior work by quantifying boththe increase in annual medical ex-penditures and the dollar value asso-ciated with increased absenteeism

46 American Journal of Health Promotion

separately for four categories ofobese men and women employed fulltime in the labor market: BMI 5 25to 29.9 kg/m2 (overweight), BMI 530 to 34.9 kg/m2 (grade-I obesity),BMI 5 35 to 39.9 kg/m2 (grade-IIobesity), and BMI 5 40 kg/m2 andgreater (grade-III obesity) (BMI iscalculated as weight in kilograms di-vided by height in meters squared).7The prevalence of grade-III obesity isincreasing at a much faster rate thanobesity in general,8 and because theadverse health consequences of obe-sity increase with increasing weight, itis likely that the adverse financialconsequences, in terms of both in-creased medical spending and in-creased absenteeism, will show a simi-lar trend.

As a result of the rapid rise inobesity rates, employers will be in-creasingly faced with difficult deci-sions concerning whether to fundworksite wellness programs, medicalor surgical interventions, and/or dis-ease-management programs relatedto obesity. These results will assist inthat determination.

METHODS

DesignStandard econometric methods

were used to separately estimate over-weight and obesity-attributable absen-teeism and medical expenditures forindividuals employed full-time for theentire year. Absenteeism was definedas annual missed work days due toillness or injury. Medical expendi-tures included annual spending foroffice-based visits, hospital outpatientvisits, hospital emergency-room visits,hospital inpatient stays, dental visits,home health care, vision aids, othermedical equipment and services(e.g., orthopedic items, medicalequipment, and disposable supplies),and prescribed medicines.

SampleThe National Health Interview

Survey (NHIS) and the Medical Ex-penditure Panel Survey (MEPS),both nationally representative andpublicly available (with responserates of at least 60%), are the prima-ry data sets used in this analysis. TheNHIS is a cross-sectional health sur-

vey conducted by the Centers for Dis-ease Control and Prevention’s(CDC’s) National Center for HealthStatistics (NCHS). It is the principalsource of information on the healthof the civilian, noninstitutionalized,household population of the UnitedStates. The sampling plan follows amultistage area probability designthat permits the representative sam-pling of households. The MEPS, ad-ministered by the Agency for Health-care Research and Quality (AHRQ),augments NHIS by selecting a sampleof NHIS respondents, collecting dataon their health care expenditures,and linking these data with addition-al information collected from the re-spondents’ medical providers, em-ployers, and insurance providers.

For the present analysis, we re-stricted the 2001 and 2002 NHIS andthe 2000 and 2001 MEPS populationsto those 18 to 64 years old who wereemployed full time, defined as great-er than 35 hours of work per week,for the entire year. Additionally, weexcluded pregnant women and per-sons missing self-reported height andweight. The final dataset for thework-loss analysis (based on theNHIS) comprised 25,427 adults(14,179 men and 11,248 women),and the final dataset for the medicalexpenditures analysis (based on theMEPS) comprised 20,329 adults(11,849 men and 8,480 women).Both analysis datasets includedweighting variables that allow forgenerating nationally representativeestimates of the U.S. full-time em-ployed population.

MeasuresThe NHIS and the MEPS ques-

tionnaires were conducted using acomputer-assisted personal interview-er (CAPI). The CAPI questionnaireswere administered using a laptopcomputer and interviewers enteredresponses directly into the computer.In addition to sociodemographiccharacteristics (e.g., gender, age, edu-cation, and income), these question-naires collected information on eachparticipant’s height and weight (laterconverted to BMI, kg/m2) and em-ployment status (e.g., hours workedper week and occupation based onthe Standard Occupational Classifica-

tion System). NHIS participants werealso asked to respond to the follow-ing question: ‘‘During the past 12months, about how many days didyou miss work at a job or businessbecause of illness or injury (do notinclude maternity leave)?’’ The re-sponse to this question was used as ameasurement of absenteeism.

MEPS medical expenditures datawere collected from each participantat the event level (e.g., doctor visit,hospital stay) and summed to pro-duce annual estimates. MEPS alsocollected data from a sample of med-ical providers and pharmacies thatwere used by participants. These ad-ditional data were used to improvethe overall quality of MEPS expendi-ture estimates.

Work-loss AnalysisWe used the 2001 and 2002 NHIS

to predict obesity-attributable absenc-es due to illness or injury. Overall,50% of the weighted regression pop-ulation (45% of men and 58% ofwomen) reported missing at least 1day of work due to illness or injury.

Due to the discrete nature of thedependent variable (i.e., daysmissed), we used negative binomialregressions with a log link to estimatemissed work days due to illness or in-jury separately for men and women.The regressions included dichoto-mous variables that indicated eachperson’s BMI category (i.e., under-weight, overweight, grade-I obese,grade-II obese, or grade-III obese),with normal weight constituting theomitted reference category. Thisspecification allowed us to predictthe impact of excess body weight onannual missed work days. The regres-sions also controlled for other factorsexpected to influence absenteeism,including age, race/ethnicity, familysize, education, income category,hourly or salaried employee, occupa-tion, years at current job, smokingstatus, alcohol use, and any function-al limitations not self-reported asobesity related (e.g., difficulty walk-ing, standing, sitting, stooping, reach-ing, or grasping).

To estimate the dollar value ofoverweight and obesity-attributablemissed work days, we used the follow-ing procedure:


1. Use the regression output to pre-dict annual missed work days foreach employee.

2. For those with a BMI greater than25, repeat step 1, setting the di-chotomous BMI category variableto zero. Because normal weightwas the omitted reference catego-ry, this approach estimated missedwork days for the hypotheticalnormal-weight employee with de-mographic characteristics identicalto those in the respective over-weight or obesity grade.

3. Subtract missed work days predict-ed in step 1 from missed workdays predicted in step 2 to esti-mate each employee’s annualoverweight or obesity-attributablemissed work days.

4. Use detailed annual employmentincome data for 2001 (the mostcurrent data available) from theU.S. Census Bureau to assign eachemployee with overweight or obe-sity-attributable missed work days adaily compensation based on his/her occupation, gender, and race.Inflate salaried employees’ dailycompensation, but not that ofhourly employees, to include thecost of benefits using a multiplierof 1.31. Inflate all estimates to2004 dollars using the All ItemsConsumer Price Index (CPI) infla-tor from the Bureau of Labor andStatistics.

5. Multiply each employee’s predict-ed overweight- or obesity-attribut-able missed work days (from step3) by their daily compensation(from step 4) to estimate the dol-lar value of missed work days foreach employee.

6. Calculate a weighted average dol-lar value that quantifies the ex-pected annual cost of absenteeismfor overweight and each obesitygrade.

Medical-expenditures AnalysisWe used the 2000 and 2001 MEPS

to predict overweight- and obesity-at-tributable medical expenditures. Thepresence in the MEPS data of bothbody measurements and annual ex-penditures allows for a direct analysisof the relationship between BMI andmedical spending. Prior to estimatingthe model, we inflated all expendi-

tures to 2004 dollars using the Medi-cal Care CPI inflator.

We used a four-equation regres-sion approach to predict annualmedical expenditures, separately formen and women, as a function ofBMI category. (Regression results forboth the work-loss and medical-ex-penditure analyses are available fromthe authors upon request.) The four-part regression approach was pio-neered by authors of the RANDHealth Insurance Experiment and isnow commonly applied to medicalexpenditures data.10,11 This approachcontrols for several unique character-istics of the medical expenditures dis-tribution, including (1) a large per-centage of people have no expendi-tures in the year (15% in our sam-ple), (2) the distribution ofexpenditures is different for users ofinpatient services relative to thosewith positive medical expendituresbut no inpatient utilization, and (3)the distribution of expendituresamong users tends to be highlyskewed, with most observations clus-tered at low values and a few obser-vations with very high values.The model estimates predicted ex-penditures using the following func-tional form:

EXP 5 P[C 3 EXPIP

1 (1 2 C)EXPNIP] (1)

where

EXP represents predicted annualexpenditures,

P represents the predictedprobability of positive medi-cal expenditures during theyear and is estimated via alogistic regression,

C represents the conditionalprobability of positive inpa-tient expenditures givenpositive expenditures and isalso estimated via a logisticregression,

EXPIP represents ordinary leastsquares (OLS)-predictedmedical expenditures givenpositive inpatient expendi-tures during the year, and

EXPNIP represents OLS-predictedmedical expenditures givenpositive expenditures but noinpatient expenditures.

All OLS regressions are estimatedon the logged expenditure variableto adjust for the skewness in annualexpenditures (mean annual expendi-tures are significantly greater thanthe median). Logged expendituresare converted back to expendituresusing a homoscedastic smearing fac-tor. Dichotomous variables that indi-cated each person’s BMI category al-lowed us to predict the impact of ex-cess body weight on annual medicalspending for each category of over-weight/obese employees. The nor-mal-weight group constituted theomitted reference category. As withthe work-loss analysis, all regressionscontrolled for other factors expectedto influence annual medical expendi-tures that were available in the data,including age, race/ethnicity, educa-tion, income category, Census region,and smoking status. The regressionspecifications used for the medicalexpenditures analysis are nearly iden-tical to those used in a recent studythat quantified national medical ex-penditures attributable to obesity.5For both the work-loss and medical-expenditures analyses, bootstrappedstandard errors were computed usingthe nonparametric BSAMPLE proce-dure in Stata. Stratum and PSU vari-ables were incorporated into thebootstrap to account for the complexsurvey design of the MEPS andNHIS.

RESULTS

Table 1 presents the self-reportedprevalence of overweight and obesityamong full-time employees aged 18to 64 from the 2001 and 2002 NHIS,and separately from the 2000 and2001 MEPS. For men, approximately70% of the full-time employed popu-lation is classified as either over-weight (;46%) or obese (;23%).For women, approximately 53% ofthe full-time employed population isclassified as either overweight(;28%) or obese (;23%). The 20-percentage point difference betweenmen and women primarily resultsfrom a much higher self-reportedprevalence of overweight among em-ployed men, 46% vs. 28%. Inversely,nearly 50% of employed women self-report a normal BMI, whereas this


Table 1

Distribution of Self-reported Body Mass Index for Full-time EmployeesAge 18 to 64 Years Old, 2001–2002*

Percent

Normal(BMI 18.5–24.9)

Overweight(BMI 25.0–29.9)

Grade-IObese

(BMI 30.0–34.9)

Grade-IIObese

(BMI 35.0–39.9)

Grade-IIIObese

(BMI $ 40.0)

NHIS

MenWomen

3049

4728

1814

46

12

MEPS

MenWomen

3046

4528

1614

56

24

* Source: Authors’ calculations based on data from the 2001 and 2002 National Health Inter-view Surveys and the 2000 and 2001 Medical Expenditures Panel Surveys.

Figure 1

Predicted Missed Work Days for Full-time Employees by BMI Category, 2001 to 2002

figure drops to 30% for employedmen.

Figure 1 presents predicted annu-al missed work days due to illness orinjury for full-time employees strati-fied by BMI category and gender.Normal-weight men miss an averageof 3.0 days each year due to illnessor injury. Annual missed work daysfor overweight and grade-I obese

men are within half a day of this esti-mate and not statistically different.Grades-II and -III obese men, howev-er, miss approximately 2 more workdays per year than normal-weightmen (p , .05).

Normal-weight women miss an av-erage of 3.4 days each year due to ill-ness or injury. In comparison, over-weight women miss 3.9 days, .5 more

days than normal-weight women;grade-I obese women miss 5.2 days,1.8 days more than normal weightwomen; and grade-II obese womenmiss 6.4 days, 3.0 days more thannormal-weight women. For womenwith a BMI of 40 kg/m2 or greater(grade-III obese), work days misseddue to illness or injury jump to 8.2days, almost a week more of missedwork each year than normal-weightwomen. All increases in absenteeismamong women are statistically signifi-cant (p , .05).

Figure 2 presents estimated annu-al per capita medical expendituresfor full-time employees stratified byBMI category. Average annual medi-cal expenditures for those in the nor-mal-weight category are estimated tobe $1351 for men and $1956 forwomen. With the exception of over-weight men, expenditure estimatesfor subsequent BMI categories arestatistically greater than the normal-weight estimate for the correspond-ing gender (p , .05).

Table 2 presents the medical ex-penditures and dollar value of in-creased absenteeism for men and


Figure 2

Predicted Annual Per Capita Medical Expenditures for Full-time Employees Age 18 to 64 Years Old by BMI Category(2004 Dollars)

Table 2

Increased Medical Expenditures and the Dollar Value of Increased AbsenteeismAssociated With Overweight and Obesity (2004 Dollars)*

Overweight(BMI 25.0–29.9)

Grade-IObese

(BMI 30.0–34.9)

Grade-IIObese

(BMI 35.0–39.9)

Grade-IIIObese

(BMI $ 40.0)

Men

Medical ExpendituresAbsenteeism

Total

$169$6

$175

$392$70

$462

$569$643

$1212

$1,591$436

$2027

Women

Medical ExpendituresAbsenteeism

Total

$495$93

$588

$1071$302

$1,372

$1549$936

$2,485

$1359$805

$2164

* Obesity-attributable medical expenditure estimates are inflated to 2004 dollars using the Med-ical Care CPI inflator. The dollar value of increased absenteeism is inflated to 2004 using the AllItems CPI inflator.

women. For men employed full timein the labor market, the combinedannual cost of obesity, includingmedical expenditures and absentee-ism, ranges from $175 per over-weight male ($169 for increased

medical expenditures and $6 for in-creased work loss) to $2027 pergrade-III obese male ($1591 for in-creased medical expenditures and$436 for increased work loss).

For women employed full time in

the labor market, due to larger in-creases in both medical expendituresand absenteeism, obesity-attributablespending is substantially higher thanthat for similar men. Overall, the an-nual cost of excess weight among fe-male employees ranges from $588 foroverweight ($495 for medical expen-ditures and $93 for absenteeism) to$2485 for grade II obesity ($1549 formedical expenditures and $936 forabsenteeism).

DISCUSSION

This analysis focused on quantify-ing the annual per capita increase inmedical expenditures and absentee-ism associated with overweight andthree unique obesity grades amongfull-time employees. The results showthat these costs are substantial. Theyrange from $176 per year for over-weight male employees to $2485 forgrade-II obese female employees. Toput these estimates into perspective,using the MEPS obesity prevalence(see Table 1) and gender distribu-


tion (59% male) among full-time em-ployees, the costs of obesity alone(excluding overweight) at a firm with1000 employees are estimated to be$285,000 per year; approximately30% of this total results from in-creased absenteeism. Separate analy-ses (not reported) reveal that firmswith a greater percentage of womenor, on average, older employeeswould have even greater expectedcosts. Moreover, although the MEPSdata suggest that those with a BMI of401 (i.e., grade-III obesity) representonly 3% of the employed population,they are responsible for 21% of thecosts of obesity.

As noted in the Introduction, notonly is the overall prevalence of obe-sity increasing, but the proportion ofthe obese population with grades-IIand -III obesity, those who we showto be most expensive, is also increas-ing. This suggests that, without inter-vention, the costs of obesity amongfull-time employees will likely contin-ue to rise. These increases will ad-versely impact the employer and allemployees, as many firms are passingalong the increases in health insur-ance directly to employees in theform of higher premiums or optingto no longer provide health insur-ance at all. (A recent survey by theWilliam M. Mercer Consulting Com-pany found that the percentage ofemployers with 10–49 employees of-fering a health plan fell from 66% to62% in 2002, and that this percent-age is expected to decrease again in2003. The Mercer survey noted that25% of all employers, and 49% oflarge employers, say that employeespaid a larger share of health plancosts in 2003 than in 2002.)12

These results present the averagemedical and absenteeism costs asso-ciated with varying degrees of obesi-ty among full-time employees. Thisinformation will allow employers tomake more informed decisions onwhether or not to pursue obesityprevention and related activities fortheir employees. In addition tothese costs, firms would ideally liketo have additional information onthe portion of the costs that couldbe reduced through effective inter-vention strategies and the fractionof these savings that would accrue to

the firm. Firms could then comparethe potential benefits to the inter-vention costs and determine the ap-propriate level and type of invest-ment. To date, much of that infor-mation remains unavailable. Regard-less, many firms have also begun tolook for strategies to improve thehealth of their employees. The ris-ing prevalence of obesity and the re-sulting increases in medical expendi-tures and absenteeism provide addi-tional motivation for employers toseek such strategies, even in the ab-sence of a full accounting of poten-tial costs and benefits. These strate-gies may include implementingworksite wellness programs, in-creased coverage for medical or sur-gical interventions (e.g., bariatricsurgery), and/or disease-manage-ment programs related to obesity.

Our analysis has several limita-tions. First, both the MEPS andNHIS rely on self-reported heightand weight, and overweight andobese people tend to underestimatetheir weight and all persons tend tooverestimate their height.13,14 As a re-sult, our overweight- and obesity-prevalence estimates are conservative.The impact of this bias on the costestimates is unclear because some in-dividuals in the normal-weight cate-gory will actually have a BMI greaterthan 25. Second, these estimates usecross-sectional data to quantify costsof overweight and obesity; they donot measure potential savings butrather the amount that would havebeen avoided had obesity never exist-ed. Therefore, as noted above, itwould be inappropriate to assumethat these estimates represent savingsfrom effective interventions. Eventhe most effective interventions willnot be able to eliminate all of thesecosts, as some of the adverse healtheffects of obesity are not reversible.Third, we were unable to computestandard errors for the estimated val-ue of work loss in Table 2; our impu-tation of daily compensation basedon census data introduced an un-known amount of variance, and thecensus data contain sampling error.Moreover, because there is some evi-dence that obese individuals (primar-ily white women) earn lower wages,our estimates on the value of in-

creased absenteeism may be biased.15

Finally, our estimates are limited to ab-senteeism and medical expendituresand therefore do not consider otherpotential costs associated with obesity,including disability and presenteeism(i.e., reduced on-the-job productivity).As a result, our cost estimates likelyunderreport the actual costs of obesityamong full-time employees. Regardless,these results provide further evidenceof the adverse financial consequencesresulting from obesity.

SO WHAT? Implications forHealth Promotion Practitionersand Researchers

This study suggests that bothmedical expenditures and absen-teeism among full-time employeesincrease with increasing weight. Asprevious research has demonstrat-ed, not only is the overall preva-lence of obesity increasing, butthe proportion of the obese popu-lation with grades II and III obesi-ty, those who this study shows tobe most expensive, is also increas-ing. Taken together, these findingssupport the assertion that obesityis a significant contributor to ris-ing health care costs. Becausesome of these costs are shoul-dered by employers, as the preva-lence and costs of obesity increase,so too does their motivation tolook for strategies to reduce thesecosts. The prevalence and cost es-timates presented in this analysis,combined with information on theeffectiveness of specific interven-tions, will allow employers to makemore informed decisions concern-ing the best strategies for reducingobesity and related costs amongtheir employees.

References

1. Flegal K, Carroll M, Ogden C, Johnson C.Prevalence and trends in obesity among USadults, 1999–2000. JAMA. 2002;288:1723–1727.

2. Calle E, Rodriquez C, Walker-Thurmond K,Thun M. Overweight, obesity, and mortalityfrom cancer in a prospectively studied cohortof US adults. New Engl J Med. 2003;384:1625–1638.

3. Must A, Spadano J, Coakley EH, et al. Thedisease burden associated with overweightand obesity. JAMA. 1999;282:1523–1529.

4. Sturm R. The effects of obesity, smoking, anddrinking on medical problems and costs: obe-sity outranks both smoking and drinking in


its deleterious effects on health and healthcosts. Health Aff. 2002;21:245–253.

5. Finkelstein E, Fiebelkorn I, Wang G. Nationalmedical expenditures attributable to over-weight and obesity: how much, and who’spaying? Health Aff—Web Exclusive May 14,2003:219–226.

6. Thompson D, Edelsberg J, Kinsey K, Oster G.Estimated economic costs of obesity to USbusiness. Am J Health Promot. 1998;13:120–127.

7. National Heart, Lung, and Blood Institute incooperation with The National Institutes ofDiabetes and Digestive and Kidney Diseases.The Clinical Guidelines on the Identification, Eval-uation, and Treatment of Overweight and Obesity

in Adults. The Evidence Report. Bethesda, Md:National Institutes of Health. NIH publication98-4083; September 1998.

8. Sturm R. Increases in clinically severe obesityin the United States, 1986–2000. Arch Int Med.2003;163:2146–2148.

9. Grosse S. Productivity loss tables (AppendixI). In: Haddix A, Teutsch S, Corso P, eds. Pre-vention Effectiveness: A Guide to Decision Analysisand Economic Evaluation. 2nd ed. London: Ox-ford University Press; 2003:245–257.

10. Manning W. The logged dependent variable,heteroscedasticity, and the retransformationproblem. J Health Econ. 1998;17:283–295.

11. Manning W, Newhouse J, Duan N, et al.

Health insurance and the demand for medi-cal care: evidence from a randomized experi-ment. Am Econ Rev. 1987;77:251–277.

12. William M. Mercer Consulting Company.Available at: http://www.mercerhr.com/pressrelease/details.jhtml?idContent51076975. Accessed January 15, 2004.

13. Rowland ML. Self-reported weight andheight. Am J Clin Nutr. 1990;52:1125–1133.

14. Palta M, Prineas RJ, Berman R, Hannan P.Comparison of self-reported and measuredheight and weight. Am J Epidemiol. 1982;115:223–230.

15. Cawley J. The impact of obesity on wages. JHum Resources. 2004;39:451–474.

A fusion of the best of science and the best of practice —

together, to produce the greatest impact.

Stay on top of the science and art of health promotion withyour own subscription to the American Journal ofHealth Promotion.

“ The American Journal of Health Promotion provides a forum for that rare commodity— practical and intellectual exchange between researchers and practitioners. ”

Kenneth E. Warner, PhDAvedis Donabedian Distinguished University Professor of Public HealthSchool of Public Health, University of Michigan

“ The contents of the American Journal of Health Promotion are timely, relevant, andmost important, written and reviewed by the most respected researchers in our field. ”

David R. Anderson, PhDVice Programs and Technology, StayWell Health Management

Definition of Health Promotion

“Health Promotion is the science and art of helping peoplechange their lifestyle to move toward a state of optimalhealth. Optimal health is defined as a balance of physical,emotional, social, spiritual and intellectual health. Lifestylechange can be facilitated through a combination of effortsto enhance awareness, change behavior and createenvironments that support good health practices. Of thethree, supportive environments will probably have thegreatest impact in producing lasting change.”

(O’Donnell, American Journal of Health Promotion, 1989, 3(3):5.)

DIMENSIONS OFOPTIMAL HEALTH

Subscribe today...

Editor in Chief Michael P. O’Donnell, PhD, MBA, MPHAssociate Editors in Chief Bradley J. Cardinal, PhDDiane H. Morris, PhD, RDJudy D. Sheeska, PhD, RDMark G. Wilson, HSD

SECTION EDITORSInterventionsFitness

Barry A. Franklin, PhDMedical Self-Care

Donald M. Vickery, MDNutrition

Karen Glanz, PhD, MPHSmoking Control

Michael P. Eriksen, ScDWeight Control

Kelly D. Brownell, PhDStress Management

Cary Cooper, CBEMind-Body Health

Kenneth R. Pelletier, PhD, MD (hc)Social Health

Kenneth R. McLeroy, PhDSpiritual Health

Larry S. Chapman, MPHStrategiesBehavior Change

James F. Prochaska, PhDCulture Change

Daniel Stokols, PhDHealth Policy

Kenneth E. Warner, PhDApplicationsUnderserved Populations

Ronald L. Braithwaite, PhDHealth Promoting Community Design

Jo Anne L. Earp, ScDResearchData Base

David R. Anderson, PhDFinancial Analysis

Ron Z. Goetzel, PhDMethod, Issues, and Results in Evaluationand Research

Lawrence W. Green, DrPHQualitative Research

Marjorie MacDonald, BN, PhDMeasurement Issues

Shawna L. Mercer, MSc, PhD

The Art of Health PromotionLarry S. Chapman, MPH

ANNUAL SUBSCRIPTION RATES: (Good through 12/31/05)

Individual Institution

U.S. $99.95 $144.85

Canada and Mexico $108.95 $153.85

Other Countries $117.95 $162.95

CALL 800-783-9913 (U.S. ONLY) or 818-760-8520OR FIND US ON THE WEB AThttp://www.HealthPromotionJournal.com

Documents

The Costs of Obesity Among Full-time Employees