Socioeconomic costs of overweight and obesity in Korean adults.

This study was conducted to estimate the socioeconomic costs of overweight and obesity in a sample of Korean adults aged 20 yr and older in 2005. The socioeconomic costs of overweight and obesity include direct costs (inpatient care, outpatient care and medication) and indirect costs (loss of productivity due to premature deaths and inpatient care, time costs, traffic costs and nursing fees). Hypertension, diabetes mellitus, dyslipidemia, ischemic heart disease, stroke, colon cancer and osteoarthritis were selected as obesity-related diseases. The population attributable fraction (PAF) of obesity was calculated from national representative data of Korea such as the National Health Insurance Corporation (NHIC) cohort data and the 2005 Korea National Health and Nutrition Examination Survey (KNHANES) data. Direct costs of overweight and obesity were estimated at approximately U$1,081 million equivalent (men: U$497 million, women: U$584 million) and indirect costs were estimated at approximately U$706 million (men: U$527 million, women: U$178 million). The estimated total socioeconomic costs of overweight and obesity were approximately U$1,787 million (men: U$1,081 million, women: U$706 million). These total costs represented about 0.22% of the gross domestic product (GDP) and 3.7% of the national health care expenditures in 2005. We found the socioeconomic costs of overweight and obesity in Korean adults aged 20 yr and older are substantial. In order to control the socioeconomic burden attributable to overweight and obesity, effective national strategies for prevention and management of obesity should be established and implemented.

INTRODUCTION

Obesity is a state of an excess of body fat that causes increased risk of metabolic derangement (1, 2). World Health Organization (WHO) estimated that globally in 2005, approximately 1.6 billion adults were overweight and at least 400 million adults were obese. They also predicted that by 2015, approximately 2.3 billion adults would be overweight and more than 700 million would be obese. Obesity prevalence has been constantly increasing in Korea, too. According to the Korea National Health and Nutrition Examination Survey (KNHANES) data, obesity (body mass index [BMI] ≥ 25 kg/m2) prevalence in Korea increased from 26.3% (male, 25.0%; female, 27.0%) in 1998 to 31.7% (male, 35.1%; female, 28.0%) in 2005. In 1997, WHO anticipated that obesity, along with smoking, would be the most serious public health problem, which threatens the health of world population, in the 21st century (3). Must et al. (4) reported that the prevalence of obesity-related comorbidities such as type 2 diabetes mellitus, high blood pressure, and osteoarthritis, was increased along with increasing severity of overweight and obesity in adults aged 25 yr and older, who participated in the Third National Health and Nutrition Examination Survey (NHANES). The disease burden associated with obesity is not only a problem limited to Western developed countries anymore but has increased to a serious level also in Asia-Pacific countries including Korea. The Asia-Pacific Cohort Studies Collaboration reported that the population attributable fraction (PAF) of overweight and obesity ranged from 0.8%-9.2% for coronary heart disease mortality, 0.2%-2.9% for hemorrhagic stroke mortality, and 0.9%-10.2% for ischemic stroke mortality in 14 Asia-Pacific countries (5).

Increase in the prevalence of obesity and obesity-related diseases leads to the growth of socioeconomic burden. According to previous studies, patients with obesity had 25%-52% higher medical care costs than normal-weighted individuals (6-8). Medical care costs attributable to obesity was reported to account for 2.0%-7.0% of national health care expenditures in Western developed countries (9). The direct costs for obesity in the USA in 1995 were estimated to be approximately 70 billion US dollars, which exceeded estimates of the direct costs for coronary artery disease, hypertension and diabetes mellitus (10). According to a study in Sweden, the obese women had 1.5-1.9 times higher sick-leave during 1 yr compared to the general Swedish women. And, approximately 10% of the total cost of loss of productivity due to sick-leave and disability pensions in Swedish women was related to obesity and obesity-related diseases (11). Obesity also brings about premature death and disability, which lead to loss of labor productivity (12, 13). Popkin et al. (14) estimated the indirect costs of obesity due to premature death, premature disability and sick-leave in China in 2000 to be 43.6 billion US dollars (3.58% of the gross national product [GNP]).

The socioeconomic costs of obesity in Korean adults were first estimated by Jeong et al. (15) using 1998 KNHANES data. They reported that the estimated costs attributable to obesity in Korea ranged from U$170 million equivalent to $350 million (exchange rate in 1998: 1,209 Korean Won for 1 US dollar), which were considerably smaller than those in Western developed countries. However, because the prevalences of obesity and obesity-related diseases in Korea have been constantly increasing, the socioeconomic costs of obesity in Korean adults are anticipated to have increased rapidly as well. Yet, there has been no study on this issue in Korea since 1998. Thus, this study was conducted to estimate the socioeconomic costs attributable to overweight and obesity in Korean adults in 2005, in which KNHANES was most recently carried out.

MATERIALS AND METHODS

Selection of obesity-related diseases

We selected hypertension (ICD-10 code: I10), diabetes mellitus (E10-E14), dyslipidemia (E78), ischemic heart disease (I20-I25), stroke (I60-I64), breast cancer (C50), colon cancer (C18-C21) and osteoarthritis (M15-M19) as obesity-related diseases because previous studies (9) demonstrated that obesity has the etiological role on these diseases and these diseases were also investigated in 2005 KNHANES. However, breast cancer was excluded from obesity-related diseases because relative risk of obesity on breast cancer was not high.

Calculation of population attributable fraction (PAF)

We used the PAF of obesity in order to allocate the proportion attributable to obesity in the total costs of obesity-related diseases. The PAF of obesity is calculated using the formula; PAF = P × (RR-1)/[P × (RR-1) + 1], where P is the prevalence of overweight and obesity and RR is the relative risk of overweight and obesity on obesity-related diseases. That is, the PAF of obesity is explained by the prevalence of obesity and the relative risk of obesity. The RRs and PAFs of obesity on obesity-related diseases in male and female were presented in Table 1 and 2.

Table 1

RR and PAF on obesity-related diseases of obesity status (men)

PAF, population attributable fraction; RR, relative risk; CI, confidence interval.

Table 2

RR and PAF on obesity-related diseases of obesity status (women)

PAF, population attributable fraction; RR, relative risk; CI, confidence interval.

Prevalence of obesity

The prevalence of obesity in Korean adults aged 20 yr and older was obtained from the raw data of 2005 KNHANES, to which was applied the weight value. We defined BMI of 18.5-22.9 kg/m2 as normal weight, 23-24.9 kg/m2 as overweight, 25-29.9 kg/m2 as obesity I and over 30 kg/m2s as obesity II. These BMI cut-offs were recommended for Asians by WHO (16). In this study, 27.4% of men and 22.0% of women were diagnosed as overweight, 31.5% of men and 24.6% of women as obese I, 3.6% of men and 3.4% of women as obese II.

PAFs of overweight, obesity I and obesity II were calculated separately and the sum of calculated PAFs was presented as the total PAF of obesity on obesity-related diseases.

Relative risk

The data source for calculation of the RR was the National Health Insurance Corporation (NHIC) cohort data. NHIC provides biennial health examinations to the entire Korean population aged 40 yr and older. The NHIC cohort was composed of the participants of health examinations of NHIC performed in 2000 who repeatedly participated in health examinations in 2002 or 2004. Subjects, who did not have any previous history of obesity-related diseases and were judged as healthy men in 2000, were only enrolled in this study. Subjects who had problem in qualification as insured persons or who had some errors in computerized data, were excluded additionally. The payment data to them by NHIC from 1999 to 2006 was obtained. The total number of the participants of health examinations of NHIC was 5,099,737. Among them, 3,026,483 and 2,958,706 individuals were again undergone health examinations in 2002 and in 2004, respectively. Finally, 1,910,194 individuals were included in this study (men, 65.4%; women, 34.5%).

The RRs of overweight, obesity I and obesity II were estimated using the Cox proportional hazard model upon the first occurrence of payment for outpatient care, inpatient care and either, respectively. Each RR was adjusted for age, smoking status, alcohol consumption, frequency of meat intake, exercise, family medical history, household income, diabetes mellitus, hypertension, hypercholesterolemia, and so on. If any RRs of obesity on obesity-related diseases were interpreted as statistically insignificant, they were not included in calculation of PAF.

Estimation of costs

The socioeconomic costs of obesity can include direct costs, indirect costs and intangible costs. Intangible costs such as pain, suffering and decreased quality of life, which are very important problems to patients, are omitted in most studies on cost of illness because it is difficult to quantify these costs in monetary terms. These costs were not included in this study. Direct costs include hospitalization, outpatient visits and medication for the purpose of prevention and treatment of obesity and obesity-related diseases. However, it is quite difficult to estimate precisely the costs for management of obesity itself, which is not covered by the national health insurance in Korea and is done in outside of the formal health care system. Costs for management of obesity-related diseases were only included as direct costs in this study. Loss of productivity due to premature deaths and admission, time costs, traffic costs and nursing fees were measured as indirect costs in this study.

Direct costs

Direct costs were composed of costs for inpatient care, outpatient care and medication. They were calculated from the payment data of NHIC, 2005 using formula 1. Costs, not insured by NHIC, were not included in these data. So, the proportion of not-insured, self-imposed costs in 2005, which was reported in an investigation (17), was reflected in this study. However, the proportions of not-insured, self-imposed costs on all obesity-related diseases were not reported in the investigation. Those were substituted with the average proportion of not-insured, self-imposed costs on all the investigated diseases. The proportions of not-insured, self-imposed costs for medication were totally applied by the average proportion of not-insured, self-imposed costs for the all investigated medication because those were not investigated separately for each disease.

Formula 1) Direct costs

= (ΣHij × PAFij)/(1-αi)

i = 1,2,....,nth disease

j = 1 if male, 2 if female

Hij = Costs for inpatient care, outpatient care or medication to treat 'i' disease in 'j' sex

PAFij = population attributable fraction of obesity on 'i' disease in 'j' sex

α = the proportion of not-insured, self-imposed costs on the 'i' disease

Indirect costs: Loss of productivity due to premature deaths

Costs by premature deaths due to obesity-related diseases were estimated by calculation of present value of lost future earning via formula 2. According to the data of the Korea National Statistical Office, total 56,633 individuals died of all obesity-related diseases in adults aged 20 yr and older in 2005 (18). The number of death attributable to obesity was calculated through multiplication the number of deaths due to each obesity-related disease by total PAF for each disease. Future earnings were calculated within wages and future earnings after 2005 were assumed to be the same as 2005. In addition, future earnings were discounted to the present values by the rate of 6%. The discount rate of 6% has been used generally in healthcare project (19). The average monthly wage income of Korean adults aged 20 yr and older in 2005 was U$ 2,084 men and U$1,383 for women (exchange rate in 2005: 1,013 Korean Won for 1 US dollar). Labor force participation rate and employment rate were 80.3% and 96.1% for men and 53.1% and 96.7% for women, respectively (20).

Formula 2) Loss of productivity due to premature deaths

= (Ij × ΣPAMij × Pj × Ej)/(1 + r)n

Ij = Average annual wage income of 'j' sex

PAMij = Number of death attributable to obesity via 'i' disease in 'j' sex

Pj = Labor force participation rate of 'j' sex

Ej = Employment rate of 'j' sex

r =The discount rate

n = Expected years of life lost

Indirect costs: Loss of productivity due to inpatient care

Loss of productivity due to hospitalization was calculated with days of inpatient care due to each obesity-related disease and the average daily wage income via formula 3. Total number of days of inpatient care due to all obesity-related diseases in adults aged 20 yr and older was approximately 7.88 million days in 2005. The average daily wage income of Korean adults aged 20 yr and older in 2005 was U$ 83 equivalent for men and U$55 for women.

Formula 3) Loss of productivity due to inpatient care

= Ijd × ΣNij × Dij × PAFij × Pj × Ej

Ijd = Average daily wage income of 'j' sex

Nij = Total number of case of inpatient care with 'i' disease in 'j' sex

Dij = Average days per case of inpatient care with 'i' disease in 'j' sex

Indirect costs: Time costs

Time costs were calculated with time spent for hospital admission and outpatient treatment and the average wage income per minute via formula 4. Time spent for hospital admission for each disease in each sex was investigated in 2005 KNHANES. Time spent for outpatient treatment included time spent for visiting physician, waiting time for outpatient treatment and treating time. In 2005 KNHANES, time spent for visiting physician was investigated, but the others were not investigated. So, waiting time for outpatient treatment was substituted with data investigated in 1998 KNHANES and treating time was assumed to be 3 min. The average wage income per minunte of Korean adults aged 20 yr and older in 2005 was U$ 0.17 for men and U$ 0.12 for women.

Formula 4) Time costs

= [(MINinij × ΣNij) + (MINonij × ΣOij)] × Ijm× PAFij × Pj × Ej

MINinij = minutes spent per hospital admission with 'i' disease in 'j' sex

MINonij = minutes spent per outpatient treatment with 'i' dis ease in 'j' sex

Oij = Total number of case of outpatient care with 'i' disease in 'j' sex

Ijm = Average wage income per minute of 'j' sex

Indirect costs: Traffic costs

The traffic costs included two-way traffic costs of outpatients, inpatients and caregivers of inpatients and were calculated via formula 5. In 2005 KNHANES, two-way traffic costs of outpatients and inpatients were investigated, but the other was not investigated. So, those of caregivers of inpatients was substituted with adjusted values, investigated in 1995 Korea National Health Examination & Health Behavior Survey, by 1.7, the increase rate of traffic fee during 10 yr.

Formula 5) Traffic costs

= (TCon × ΣOij) × PAFij + [(TCin + TCcg) × ΣNij] × PAFij

TCon = Average two-way transportation costs of outpatients

TCin = Average two-way transportation costs of inpatients

TCcg = Average two-way transportation costs of caregivers of inpatients

Indirect costs: Nursing fees

The nursing fees were investigated in 2005 KNHANES and were calculated via formula 6. It was investigated in 2005 KNHANES that nursing fees were paid only in colon cancer (U$ 340) in men, and stroke (U$ 877), colon cancer (U$ 237), and osteoarthritis (U$ 494) in women.

Formula 6) Nursing fees

= ΣNij × NCij × PAFij

NCij = Average nursing costs per case of inpatient care with 'i' disease in 'j' sex

RESULTS

Costs for inpatient care attributable to overweight and obesity in Korean adults aged 20 yr and older were estimated (Table 3) at about U$265 million equivalent (men, U$108 million; women, U$157 million). Inpatient care cost for ischemic heart disease occupied the largest proportion in these costs in men and inpatient care cost for osteoarthritis did in women. The estimated costs for outpatient care attributable to overweight and obesity were approximately U$300 million equivalent (men, U$133 million; women, U$167 million). Costs for medication attributable to overweight and obesity were estimated to be approximately U$516 million (men, U$256 million; women, U$260 million). Hypertension made the largest contribution to costs for both outpatient care cost and medication attributable to overweight and obesity in both sexes. The total direct costs of overweight and obesity were estimated at about U$1,081 million (men, U$497 million; women, U$584 million). The direct costs of overweight and obesity due to hypertension were the highest in both sexes. Costs for medication attributable to overweight and obesity were larger than costs for inpatients care or outpatient care.

Table 3

Direct costs of obesity (unit: U$1,000 equivalent)

HTN, hypertension; DM, diabetes mellitus; IHD, ischemic heart disease; OA, osteoarthritis.

Indirect costs

Loss of productivity due to premature deaths attributable to overweight and obesity was estimated (Table 4) at approximately U$444 million (men, U$374 million; women, U$70 million). Diabetes mellitus were the most important pathway of costs due to obestiy-attributed premature deaths in both sexes. The estimated amount of loss of productivity due to hospitalization attributable to overweight and obesity obesity-related diseases was approximately U$74 million (men, U$44 million; women, U$29 million). Hospitalization due to diabetes mellitus was the most important cause of these costs in men and hospitalization due to stroke was in women. Time costs attributable to obesity was estimated to be about U$70 million (men, U$46 million; women, U$24 million) and transportaion costs attributable to overweight and obesity were estimated to be approximately U$103 million (men, U$62 million; women, U$41 million). Hypertension made the largest contribution to time costs of overweight and obestiy and diabetes mellitus most largely contributed to transportation costs of overweight and obesity. Nursing fees attributable to overweight and obesity were estimated at about U$16 million (men, U$2 million; women, U$14 million). The total indirect costs of overweight and obesity were estimated to be approximately U$706 million (men, U$527 million; women, U$178 million). The indirect costs of overweight and obesity due to diabetes mellitus were the highest in both sexes. The indirect costs of overweight and obesity were larger in men than in women.

Table 4

Indirect costs of obesity (unit: U$1,000 equivalent)

HTN, hypertension; DM, diabetes mellitus; IHD, ischemic heart disease; OA, osteoarthritis.

Total socioeconomic costs

The estimated total socioeconomic costs of overweight and obesity in Korean adults aged 20 yr and older were approximately U$1,787 million equivalent (men, U$1,081 million; women, U$706 million). Costs for diabetes mellitus occupied the largest proportion in the total costs attributable to overweight and obesity in men and costs for hypertension did in women (Table 5). The estimated total socioeconomic costs of obesity in Korean adults aged 20 yr and older were approximately U$1,306 million and those of overweight were approximately U$481 million (Table 6).

Table 5

Total socioeconomic costs of obesity (unit: U$1,000 equivalent)

HTN, hypertension; DM, diabetes mellitus; IHD, ischemic heart disease; OA, osteoarthritis.

Table 6

Total socioeconomic costs of obesity by obesity grade (unit: U$1,000 equivalent)

Overweight, 23-24.9 kg/m2; Obesity I, 25-29.9 kg/m2; Obesity II, ≥ 30 kg/m2.

DISCUSSION

This study estimated the socioeconomic costs of overweight and obesity in Korean adults aged 20 yr and older in 2005. The estimated total costs were approximately U$1.8 billion equivalent (direct costs: U$1.1 billion, indirect costs: U$0.7 billion). These total costs represented about 0.22% of the gross domestic product (GDP) and 3.7% of the national health care expenditures in 2005. Diabetes mellitus and hypertension were the two main contributors to the socioeconomic costs of overweight and obesity.

A previous study (15) on the same issue presented that the socioeconomic cost of overweight and obesity (BMI ≥ 23 kg/m2) in Korean adults in 1998 were approximately U$350 million equivalent (0.094% of GDP, 1.88% of the total national health care expenditures). The socioeconomic costs of overweight and obesity has increased 4-5 times during these 7 yr. Even if the growth of GDP and the national health care expenditures in Korea was considered, increase of these costs was more than 2 times. It is thought that the rapid increase in the prevalence of obesity in Korea has led to the increase of the socioeconomic costs attributable to obesity.

According to a previous study (9), obesity accounts for 5.5%-7.0% of national health care expenditures in the USA and 2.0%-3.5% in other Western developed countries. The estimated medical-care costs attributable to obesity were approximately 2.3% of the national health-care expenditures, which were similar to those in other Western developed countries than the USA. These results imply that the economic burden attributable to obesity in Korea is getting bigger like Western countries.

The socioeconomic costs of obesity can be classified into tangible costs and intangible costs. Intangible costs are very important problems to patients such as pain, suffering, and anxiety and are usually measured in terms of quality of life. Although there are several studies (21-23) that estimated intangible costs of diseases, these costs are omitted in most studies on cost of illness because it is difficult to quantify these costs in monetary terms. For estimation of intangible costs, organized and meticulous investigation on the quality of life of study subjects should be planned. However, because this study was made of secondary data, intangible costs could not be included in this study.

The medical care costs attributable to obesity can be classified into costs for management of obesity-related diseases and costs for management of obesity itself. And, these costs can be divided into costs expended within the formal health care system and costs expended in outside of the formal health care system, respectively. Costs for management of obesity itself is rapidly increasing in Korea. The amount of sales of anti-obesity drugs has increased from U$60 million equivalent in 2005 to U$81 million in 2007 (exchange rate in 2007, 938 Korean Won for 1 US dollar) in Korea (22). However, management of obesity is not covered by the national health insurance in Korea and a lot of Koreans utilize services for weight loss in outside of the formal health care system. Therefore, data on the costs for management of obesity itself could not be obtained. Besides the diseases defined as obesity-related diseases in this study, various diseases including gallbladder disease and endometrial cancer can be considered as obesity-related diseases. However, only diseases investigated in 2005 KNHANES were selected as obesity-related diseases in this study. These limitations of study methods made the socioeconomic cost of overweight and obesity underestimated in this study. Therefore, it is thought that the real costs of obesity may be superior to U$1.8 billion equivalent, the estimated cost of obesty of this study. That PAF of obesity was calculated from national representative data such as NHIC cohort data and 2005 KNHANES data, is one of strengths of this study. The PAF of obesity calculated in this study can be utilized in researches on the relationship between obesity and obesity-related diseases in Asian including Korean.

We found the socioeconomic costs of overweight and obesity in Korean adults aged 20 yr and older were substantial. Also, this study showed the socioeconomic burden of obesity (U$1,306 million equivalent) was much bigger than that of overweight (U$481 million). This study might be a good evidence for benefits of prevention and management of obesity. In order to control the socioeconomic burden attributable to overweight and obesity, effective national strategies for prevention and management of obesity should be established and implemented. The results of this study are expected to make a contribution to plan and execution of obesity-related public health policy.