The US Government's National Health and Nutrition Examination Surveys (NHANES) 2003-4 and 2005-6 were used to calculate stature and weight data for the different age groups. The combined sample consisted of 3850 adult men and 4401 adult women. The raw data were weighted by the NHANES weighting factor (wtmec2yr/2) to make them representative of the US population.
The sampling quality makes this the definite record of stature and weight in the USA. (Clothing and other surveys are more prone to self-selection problems and typically show the population to be taller and lighter.)
Incomes: there is a tendency for stature and weight to vary with income, higher income is associated with taller stature among both sexes and with lower weight among women. Means were calculated from NHANES 2003-6, using the income variable to split the dataset at the 35th and 75th percentiles (the most extreme percentiles where sample size was still satisfactory). For stature just the overall mean was used, there were no consistent trends by age. The same was found for female weight. Among men weight was less among low income men in working years, especially in the range 35-55. We may speculate that this is because of their physical activity. Because of small cell sizes three age ranges were calculated: 35-55, other ages, and all ages 18-64. The differences were applied to PeopleSize age ranges so that 18-25's had 18-64 ratios, i.e. a moderate age effect, while 25-55 has the 35-55 maximum ratio.