The sugar-fat relationship revisited: differences in consumption between men and women of varying BMI

To assess the relationship of dietary fat and sugar consumption in men and women with different body mass indices (BMI). To determine the actual food sources for sugar intake, comparing differences between men and women across BMI groups. The effect of excluding individuals with low energy intakes (that is, invalid data) on these relationships was also incorporated in the analysis.

Subjects for this analysis were those individuals who participated in the 1986-1987 Dietary and Nutrition Survey of British Adults (DNSBA).

In the DNSBA, dietary intake was assessed using seven-day weighed food records, providing estimates of dietary fat and sugar intake. From the DNSBA database food records, sources of sugar intake were classified into five sugar containing food groups (high fat sweet products, fruits, dairy products, sugar products (excluding soft drinks) and sugar products (including soft drinks)). BMI was calculated from the measurement of height and body weight.

A positive relationship between BMI and dietary fat intake was found for men, both when fat was expressed as a percentage of energy and in absolute terms (g/d). This relationship was only replicated for women when intake was expressed in absolute terms. A negative relationship was found between sugar intake (as a percentage of energy) and BMI in men, but not women. Expressing sugar consumption in absolute terms did not produce a statistically significant relationship with BMI for either men or women. In women the only sugar source associated with BMI was high fat sweet products (for example, cakes, biscuits, chocolate), where higher intakes were related to higher BMIs. The reverse relationship was found for men. In men, BMI was also negatively related to the intake of sugar products (for example, table sugar, preserves, sugar confectionery), both when soft drinks were included and excluded. The inclusion of low energy reporters (LER) in the analysis altered the relationships between nutrients and BMI, particularly among women. The association between overall fat intake (g/d) and BMI was weakened, while the negative relationship with sugar intake was strengthened. In the case of women, the inclusion of LER completely reversed the relationship between consumption of high fat sweet foods (cakes, biscuits, chocolate) and BMI (due to the reduced reporting of these products by obese women). Fewer alterations in the relationships between BMI and the sources of sugar consumed were observed in men than in women when LER were included in the analysis.

The relationships between dietary fat, sugar and BMI are different in men and women, and are dependent on the inclusion of LER, particularly in women. The results suggest that among women the consumption of high fat sweet products may be a factor in understanding obesity. Furthermore, the observation of high consumption of these foods among obese women is consistent with measured preferences for these high fat sweet foods. The altered representation of the data created by LER appears to distort the relationship between sugar, fat and the degree of obesity in men and women.