The prevalence of obesity (body mass index ≥30.0 kg/m2
) in U.S. adults was derived from the 1960–62, 1971–74, 1976–80, 1988–94, 1999–2000, 2001–02, 2003–04 and 2005–06 U.S. National Health and Nutrition Examination Surveys (NHANES) 
. NHANES uses a complex, multistage, probability sampling design to select participants who are representative of the civilian, non-institutionalized U.S. population. Sample weights are assigned to each individual to represent the U.S. population and allow for the development of national prevalence figures. The prevalence of obesity for each examination period was age-adjusted by the direct method to the year 2000 U.S. Bureau of Census estimates. Mean body weight for the 1960–62, 1971–74, 1976–80, 1988–94, 1999–2002 and 2003–2006 NHANES surveys for the 40–50 year old age group were examined by gender 
. We chose this age group because it has the highest percent of employed individuals for both men and women 
. The National Center for Health Statistics (NCHS) ethics review board approved the original survey protocols, and informed consent was obtained for all NHANES participants.
Employment data were derived from the Current Employment Statistics (CES) program for the years 1960 to 2008. The CES is a monthly survey of businesses and government agencies conducted by State employment security agencies in cooperation with the Bureau of Labor Statistics. The survey provides employment, hours and earnings estimates based on payroll records of business establishments dating back to 1939, but is limited to nonagricultural industries. Occupations are broadly categorized as goods-producing or service providing. Goods-producing further sub-categorized into mining-logging, construction and manufacturing, while service-providing occupations are further divided into the categories of trade (whole sale and retail), transportation/utilities, information, financial services, professional/business services, education/health services, leisure/hospitality and other. Agricultural employment data were derived from the Current Population Survey (CPS) which is a monthly survey of households conducted by the Bureau of Census for the Bureau of Labor Statistics. The CPS provides a comprehensive body of data on the labor force including employment, unemployment and persons not in the labor force. While CPS has agricultural employment data dating back to1940, it only has detailed occupation and industry data starting in 1983.
Median, minimum and maximum physical activity intensity (Metabolic Equivalents: METs) were assigned to each occupation based on previously published classification schemes 
. Industries were then categorized into physical activity intensity groups based on the median METs value. The intensity categories were sedentary (<2 METs), light (2.0–2.9 METS), and moderate (3.0–5.9 METs) 
The prevalence of individuals in specific occupations by year was calculated by dividing the number of individuals in a given occupation by total U.S. occupations in private industry from CES combined with total agricultural occupations from the CPS for that year. The prevalence of individuals in sedentary, light, and moderate intensity category industries was calculated for each year. Mean occupational-related METs for each year were calculated as follows: the number of individuals employed in each occupation was multiplied by the median METs for the occupation with the product for each occupation added together and the sum divided by the total number of employed individuals for that year. To estimate daily occupation-related physical activity energy expenditure we assumed an 8 hour work day, used the mean weight from 1960–1962 NHANES survey (64.9 kg for women and 76.9 kg for men) and used the mean MET value for each year with the formula: daily caloric expenditure
hours worked×mean MET value×weight (kg) 
Energy Balance Differential Equation Model
Theoretical predictions of weight gain as a result of changes in energy expenditures were computed using a validated energy balance differential equation model for weight change 
. The model is derived from the first law of thermodynamics and predicts weight change resulting from changes in energy intake and energy expenditures. The predictions account for weight dependent changes in energy expenditure through specific formulations of weight dependent terms for physical activity and resting metabolic rate.
In order to compare model predictions to the NHANES data, we determined the average weight gain in kg per kcal decrease in energy expenditure resulting from the trends in occupational activities. Specifically, after decreasing energy expenditures by increments of 50, 100, and 150 kcal/d to represent decreased work-related physical activity, the model was simulated until steady state was reached. The predicted change in weight per decrease in kcal/d energy expenditure was determined by averaging the three ratios of the change in steady state weight per 50 kcal/d. From the model men gain 0.090 kg for every kcal of energy expenditure decrease and women gain 0.092 kg for every kcal of energy expenditure decrease.