Reported energy intake from different sources, sociodemographic and anthropometrical characteristics of the study population are given in table . The decrease in waist-hip ratio was mainly due to a marked increase in hip circumference in both men and women. Smoking became less common, reported energy intake from saturated fatty acids decreased, and intake of alcohol increased. Higher education became more common.
Characteristics of study population, 1986 vs. 1999
A complete list of food items that were covered by questionnaires is given in the appendix (additional file 1
: List of items on food-frequency questionnaires). Out of these, 15 items are shown that were associated with the 10 largest differences in distribution of body fat in either men or women.
Table summarizes time trends in food consumption associated with the largest differences of waist- and hip circumference for women. Increased use of (vegetable) oil and pasta as well as reduced consumption of fruit creams and 3% fat milk were all associated with reduction of waist circumference. Growing popularity of hamburgers, French fried potatoes and soft drinks were associated with an increase of waist circumference. Increased hip circumference was associated with higher consumption of pasta, vegetable oil as well as cream and 1.5% milk. Time trends for hamburgers and French fried potatoes went along with minor reductions of hip circumference. Adjustment for other lifestyle-factors attenuated the net effect of time trends in reported food consumption but, did not alter their directions. Only the protective effects of lower consumption of 3% milk and added sugar were reduced markedly. The net effect of increased consumption of 4% beer was reversed.
Food items associated with largest estimated effect on mean waist- and hip-circumference in women.
Figure illustrates associations of food items with differences in both average hip- and waist-circumference in women. Time trends for vegetable oil, pasta, fruit creams and cream were associated with risk-lowering anthropometric time trends, whereas trends for hamburgers and French fried potatoes correlated with risk-increasing trends.
Figure 2 Estimated effect of time trends in reported food intake 1986–1999 on average waist- and hip-circumference in women. 10 items with the largest estimated effect on distribution of body fat in women. Sort order is the sum of effects from largest (more ...)
In men, time trends for vegetable oil, pasta and milk were associated with both, largest increase of hip-circumference and largest reduction of waist-circumference (Table , Figure ). Increased use of hamburgers and potato chips were associated with an increase of average waist circumference but also a positive effect on hip circumference. Only rising consumption of 4% beer was associated with both, HC decrease and WC increase.
Food items associated with largest estimated effect on mean waist- and hip-circumference in men.
Figure 3 Estimated effect of time trends in reported food intake 1986–1999 on average waist- and hip-circumference in men. 10 items with the largest estimated effect on distribution of body fat in men. Sort order is the sum of effects from largest reduction (more ...)
After adjustment for lifestyle-variables (Table , model 2) the waist reducing effect of time trends in pasta consumption disappeared whereas the effect of increased consumption of wine was reversed. The negative net-effect of French fried potatoes could also be explained by associated lifestyle.
In general, waist circumference was more responsive in women whereas hip and waist circumferences were equally affected in men.