How Important is Fasting Status for Interpreting Serum Lipid and Hormone Measurements?

Purpose Fasting is thought to influence various serum lipid measurements and certain serum hormones, but to date these associations have not been examined comprehensively. Methods We obtained serum measures and fasting status for 1255 men (ages 30–79 y) in the Boston Area Community Health survey. As in clinical practice, fasting was based on self-report. Subjects who had consumed food or beverages within 8 hr of sampling were designated non-fasting (n=1034); remaining subjects were considered fasting (n=221). Analyses compared geometric mean lipid (total cholesterol, triglycerides (TGs), high density lipoprotein cholesterol, and apolipoproteins A1 and B) and hormone (testosterone, cortisol, sex hormone-binding globulin, dehydroepiandrosterone (DHEA), DHEA sulfate, luteinizing hormone, follicle-stimulating hormone, prolactin, dehydrotestosterone, estradiol, and estrone) levels in fasting vs. non-fasting subjects. Initial models adjusted for age and ethnicity; models including cholesterol medication, body mass index, and hours since waking were also considered. Results Even after covariate adjustment, most hormones differed by fasting status, particularly cortisol (13%, p<0.001) and DHEA (15%, p=0.002). Of the lipid measures, only TGs were affected. Hours since waking strongly predicted fasting status and influenced the strength of associations; initial models indicated 16% lower TGs in fasting vs. nonfasting subjects, but controlling for time of waking reduced this to 10%. Conclusion These results have implications for the costs, design, execution, and interpretation of field studies, particularly where serum hormones are considered. If fasting blood cannot be obtained, fasting status and time of waking should be collected for consideration in analyses.