gms | German Medical Science

Background: Recently it was shown that postprandial blood glucose measurements were comparable to fasting blood glucose levels. The measurement of blood lipids in a fasting status has been under question. Here we analyzed the effect of fasting duration on levels of blood triglycerides (Trig).

Materials and methods: The German Metabolic and Cardiovascular Risk Project (GEMCAS) Study was a nation-wide cross-sectional study conducted for two weeks in October 2005. The study included 35,869 participants between 18-99 years. The present analysis included 29,221 participants (59.8% females) after excluding missing data on fasting duration. We built groups based on fasting duration, which ranged from 0 to ≥12h (t0-t≥12). Because of the skewness of the Trig data we used log transformation before performing the analysis and therefore report the corresponding geometric means and their 95% confidence interval (95%-CI) for the back transformed Trig values. Since fasting duration is dependent on age we further standardized for age according to the German population (2004).

Results: The geometric mean of the reference group (t≥12) was 131.4 (95%-CI: 106.4; 162.3) for males and 105.4 (95%-CI: 88.8; 124.8) for females. Trig values between t0 to t4 were higher [highest difference at t3 (+22.3 mg/dL)] than the reference Trig values, whereas t5 to t7 and t10 to t11 were comparable to the reference. In general, we observed that the variation of the measured Trig values within each group exceeded the difference of Trig values from the reference group. Covariance analysis including age, physical activity, day time of blood withdrawal, lipid lowering agents and intake of alcohol did not change the results. However our analysis revealed that further relevant influencing factors for Trig such as physical activity and intake of lipid lowering agents should be taken into account, when measuring blood triglycerides.

Conclusion: Our study shows only small effects of fasting duration on blood lipid values latest after four hours of fasting. This challenges the necessity of measurements of fasting blood lipids, at least in epidemiological studies, which is in line with current findings [1].