Published in:
Public Health Nutrition, 8 (2005) 8, Seite 1322-1327
Language:
English
DOI:
10.1079/phn2005750
ISSN:
1368-9800;
1475-2727
Origination:
Footnote:
Description:
AbstractObjectiveTo describe the consequences of systematic reporting bias by the obese for diet–disease relationships.DesignThe present report used 24-hour urinary nitrogen and estimates of 24-hour energy expenditure to assess error in diet reporting, and examined the consequence of accounting for this error for associations between dietary fat intake and serum low-density lipoprotein (LDL)-cholesterol.SettingSub-study to the Danish MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) project, carried out in 1987–1988.SubjectsA random sub-sample of the adult Danish male population (n = 152).ResultsCorrecting dietary fat for underreporting error weakened, rather than strengthened, the association between dietary fat intake and LDL-cholesterol by reducing the slope of the regression from β = 3.4, P = 0.02 to β = 2.7, P = 0.04.ConclusionThis example illustrates that systematic underreporting of dietary fat by high-risk groups such as the obese may produce an overestimated association. These results imply that previous epidemiological studies showing a positive association between percentage of energy from fat and other health outcomes, e.g. cancer and heart disease, may have overestimated the negative effects of a high-fat diet. If we were able to correctly assess dietary fat intake in general populations, recommendations for fat intake may be more liberal than the 30% suggested today. Improved assessment of fat intake in epidemiological studies is necessary for future development of evidence-based recommendations for diet and health.