Published online December 1, 2005
PEDIATRICS
Vol. 116
No. 6
December 2005, pp.
1612-1613
(doi:10.1542/peds.2005-2269)
Inappropriate Interpretation of the Odds Ratio: Oddly Not That Uncommon
Dewesh Agrawal, MD
Division of Emergency Medicine,
Children's National Medical Center,
Washington, DC 20010
To the Editor.—
I applaud the work of Must et al1 for clarifying the effects of premenarcheal weight and maturational timing on adult weight status. However, the interpretation offered of the odds ratio of 7.7 from their Table 4 is somewhat misleading: "Girls who were overweight before menarche were 7.7 times more likely to be overweight as adults."1
To understand why this interpretation is inappropriate, one must first understand the difference between a risk ratio (relative risk [RR]) and an odds ratio (OR). An RR is the ratio of the risk of an event among an exposed population to the risk among the unexposed. An OR is the ratio of the odds of an event in an exposed group to the odds of the same event in a group that is not exposed.2 RRs can be interpreted literally as increased or decreased likelihood of an event between exposed and unexposed populations. In other words, an RR of 7.7 can be interpreted as "the event was 7.7 times more likely to occur in the exposed population than in the unexposed population." However, ORs are difficult to comprehend directly and should not be interpreted (as commonly done) as being equivalent to the RR.3 Thus, the OR of 7.7 from the Must et al study cannot be interpreted as "7.7 times more likely." It would have been more appropriate for the authors to state instead that "girls who were overweight before menarche were more likely to be overweight as adults (OR: 7.7; 95% confidence interval: 2.3–25.8)."
If the OR is interpreted inappropriately as an RR (as was done in this study), it will always overstate the effect size. In fact, the higher the event rate, the more the divergence between the OR and the RR.3 In other works, the higher the proportion is of premenarcheal overweight girls that were overweight as adults, the more that relative risk will be less than 7.7. Thus, the RR in this study is likely to be significantly less than 7.7.
REFERENCES
- Must A, Naumova EN, Phillips SM, Blum M, Dawson-Hughes B, Rand WM. Childhood overweight and maturational timing in the development of adult overweight and fatness: the Newton Girls Study and its follow-up.
Pediatrics. 2005;116
:620
–627[Abstract/Free Full Text]
- Guyant G, Rennie D.
Users' Guide to the Medical Literature: A Manual for Evidence-Based Clinical Practice. Chicago, IL: AMA Press; 2002
- Davies HTO, Crombie IK, Tavakoli M. When can odds ratios mislead?
BMJ. 1998;316
:989
–991[Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2005 by the American Academy of Pediatrics
