Obesity is a major public health problem and as such has been linked to higher medical costs¹. In 2005, the World Health Organization estimated that almost 400 million people were obese and this number was projected to double over the next 10 years². For example, in Australia the cost of obesity in 2005 was AUD $3.7 billion, of which one third were direct costs to the government¹. However, there is a lack of information between studies to show how and why excess costs were accrued in obese populations³. The present study aimed to explore the impact of obesity on the cost of disease management in people with or at high risk of atherothrombotic disease and to explore the causes (excess costs) of any difference between obese, overweight and normal weight⁴.

Cost of Obesity among subjects with or at high risk of atherothrombosis

This study was conducted as part of a nationwide prospective Australian Reduction of Atherothrombosis for Continued Health (REACH) Registry. The health-care costs of obesity were estimated for 2,819 subjects who were enrolled through primary care general practice in 2004 and who were aged ≥ 45 and had established coronary artery disease, stroke and peripheral artery disease, or ≥ 3 risk factors (hypertension, hypercholesterolemia, smoking, and diabetes). Data was collected on an internationally standardized case report form. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared kg/m². Subjects were considered to be normal weight if their BMI ranged from 18.5 to < 25.0 kg/m², overweight from 25.0 to < 30.0 kg/m² and obese if ≥ 30.0 kg/m². The medical histories within 12 months and baseline measure for comorbidities were measured as a part of the registry. Reasons for hospitalization and ambulatory care service over one year of follow-up were collected as part of the standardised case report form. The baseline medication usage was collected as part of general practitioners case notes and by a participant interview undertaken by a trained nurse. We assigned a unit cost for each health-care item used, based on the Australian Government reimbursed data for 2006-2007. We applied linear mixed models to estimate the association between direct medical costs and BMI categories.

Excess cost of obesity

Among the 2,819 subjects pharmaceutical costs per person increased with increasing BMI category. When adjusting for comorbidities the results showed that adjusted annual pharmaceutical costs of overweight and obese subjects were higher ($87 (p=0.004) and $144 (p<0.001), respectively). This was due to subjects in higher BMI categories receiving more pharmaceuticals than normal weight participants. The same relationship was not observed across BMI categories in annual ambulatory care costs and annual hospital costs.

Implications

There are several potential explanations for the greater use of pharmaceuticals among subjects with higher BMI. A previous study suggested that obese people are more likely to be treated medically rather than surgically⁵. In addition, obese people might be in need of more drugs due to a greater complication of disease. However, even if they were receiving more drugs their risk factors remained still higher. The other possible explanation is that participants with greater body weights were prescribed more drugs due to an appearance suggestive of being at higher risk. The implications of our findings are in terms of opportunity costs, given limited health-care resources governments struggle whom to subsidize first and for which illness. For example, if we consider the current rates of cardiovascular disease in Australia, for people aged over 65 years, almost 30% of them were obese. If we assume that calculated costs derived from this study are applicable to the general Australian population aged over 65 years and with or at high risk of atherothrombosis, then the excess cost to the government due to obesity is $37 million. We are questioning whether the additional pharmaceutical cost used provides an additional benefit, or whether this funding should be allocated elsewhere.