Dietary intake is believed to follow a seasonal pattern, so adjusting for seasonality in nutritional epidemiology is usually requested. The aim of this study was to assess the seasonality of energy and macronutrients intake (expressed as % of total caloric intake) using a large set of population-based studies.
Cross-sectional data from 9 population-based studies from 4 countries (3 in the Northern Hemisphere and 1 in Southern Hemisphere), with a total sample size of 44,611 subjects. Data were derived from FFQ or repeated 24 h-recalls. Dietary markers included total energy intake, protein, carbohydrates, fat, alcohol, sugars and fatty acids (saturated, mono and poly-unsaturated). Seasonality was assessed using the cosinor method stratifying on hemisphere and gender.
Most nutrients did not show a significant seasonal variation. When individual studies were considered, the number of nutrients showing significant seasonal variations varied from 5 to none in men and from 6 to none in women. Jointly, in the Northern hemisphere, significant seasonal variations were found for sugar intake in both genders and for alcohol consumption in men only; in the Southern Hemisphere, significant seasonal variations were found for fat and monounsaturated fatty acids in women. Analysis of the three consecutive periods of the Bus Santé study in Switzerland showed that the number of significant seasonal variations decreased from 5 in 1993-1999 to 1 in 2006-2012 in men, and from 6 in 1993-1999 to none in 2006-2012 in women. Seasonal variation decreased over time for most nutrients, with the exception of monounsaturated fatty acids in men.
Seasonality of energy and macronutrients intake varies considerably according to study, with no common pattern and small magnitude of variation. Its amplitude appears to be declining over time. Systematic adjustment for season might not be necessary when assessing the effects of total energy and macronutrient intake on disease.