Format

Send to:

Choose Destination
See comment in PubMed Commons below
Epidemiology. 2002 Mar;13(2):165-71.

Heterogeneity in assessing self-reports of caffeine exposure: implications for studies of health effects.

Author information

  • 1Perinatal Epidemiology Unit, Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, CT 06520-8034, USA. michael.bracken@yale.edu

Abstract

BACKGROUND:

Coffee and its metabolite caffeine are widely studied for their health effects but with inconclusive results. Caffeine is particularly difficult to assess, and therefore we explore heterogeneity of caffeine exposure.

METHODS:

We categorized caffeine exposure among 2,478 pregnant women in southern New England during 1996-2000 by the traditional laboratory-based methods of M. Bunker and M. McWilliams. A subsample was examined to ascertain caffeine levels of brewed or purchased beverages actually consumed.

RESULTS:

More than half (56.6%) of women drank coffee since becoming pregnant. Serving sizes ranged from 2 to 32 oz and are considerably larger than laboratory standards, which are typically 8-10 oz, as compared with the standard of 5 to 6 oz. Conversely, caffeine content per serving of coffee was one-third the laboratory standard, eg, 100 mg caffeine compared with 300 mg for a 10-oz cup. Tea brewed more than 3 minutes contained 42 mg caffeine as compared with the standard of 94 mg. When the amount of caffeine actually consumed was measured, one-quarter (24.8%) of subjects traditionally classified as consuming 300+ gm caffeine daily were reclassified as consuming 150-299 mg.

CONCLUSION:

Misclassification of caffeine consumption increases difficulty in identifying health effects from caffeine. Some combination of more precise consumption data and a biomarker such as paraxanthine may more precisely estimate exposure.

PMID:
11880757
[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Icon for Lippincott Williams & Wilkins
    Loading ...
    Write to the Help Desk