Misuse of Odds Ratios in Obesity Literature: An Empirical Analysis of Published Studies

Department of Health Care Organization and Policy, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA.
Obesity (Impact Factor: 3.73). 03/2012; 20(8):1726-31. DOI: 10.1038/oby.2012.71
Source: PubMed


Odds ratios (ORs) are widely used in scientific research to demonstrate the associations between outcome variables and covariates (risk factors) of interest, and are often described in language suitable for risks or probabilities, but odds and probabilities are related, not equivalent. In situations where the outcome is not rare (e.g., obesity), ORs no longer approximate the relative risk ratio (RR) and may be misinterpreted. Our study examines the extent of misinterpretation of ORs in Obesity and International Journal of Obesity. We reviewed all 2010 issues of these journals to identify all articles that presented ORs. Included articles were then primarily reviewed for correct presentation and interpretation of ORs; and secondarily reviewed for article characteristics that may have been associated with how ORs are presented and interpreted. Of the 855 articles examined, 62 (7.3%) presented ORs. ORs were presented incorrectly in 23.2% of these articles. Clinical articles were more likely to present ORs correctly than social science or basic science articles. Studies with outcome variables that had higher relative prevalence were less likely to present ORs correctly. Overall, almost one-quarter of the studies presenting ORs in two leading journals on obesity misinterpreted them. Furthermore, even when researchers present ORs correctly, the lay media may misinterpret them as relative RRs. Therefore, we suggest that when the magnitude of associations is of interest, researchers should carefully and accurately present interpretable measures of association--including RRs and risk differences--to minimize confusion and misrepresentation of research results.

Download full-text


Available from: Nir Menachemi, Feb 09, 2015
35 Reads
  • Source
    Obesity 01/2013; · 3.73 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Abstract Purpose . Existing literature shows negative relationships between gasoline price and motor vehicle crashes, particularly among teens. This paper extends that literature by evaluating the relationship between gasoline price and self-reported risky driving among teens. Design . Observational study using multivariate empirical analysis, using pooled data from the Youth Risk Behavior Survey, waves 2003-2009. Setting . Secondary data from survey administered in private and public high schools across the United States. Subjects . Students in grades 9 through 12, surveyed biennially from 2003 to 2009 (n = 58,749). Measures . Outcomes are (self-reported) driving without seatbelts, driving after consuming alcohol, and moderate physical activity (like walking or bicycling). State-level retail gasoline prices constitute the main predictor variable. Analysis . Multivariate logistic models are estimated for the full sample, as well as by gender, race/ethnicity, and age. Individual characteristics, state unemployment, and state driving policies are controlled for. Standard errors are clustered at the state level. Results are reported in form of risk differences. Results . Higher gasoline prices are negatively and significantly associated with driving without seatbelts. Associations are particularly strong for males and minorities. There are fewer statistical associations between gasoline prices and driving after drinking. Higher gasoline prices are positively associated with more moderate physical activity. Conclusion . Higher gasoline prices are associated with less risky driving behaviors among teens, and they may be associated with more active forms of transportation, like walking and bicycling. The study limitations are discussed.
    American journal of health promotion: AJHP 04/2013; 28(3). DOI:10.4278/ajhp.120926-QUAN-471 · 2.37 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: As most obese adults were not overweight as children, the prediction of adult obesity from childhood body size alone is limited. We constructed a two-way, multifactor risk assessment framework for predicting adult obesity during childhood using the Foresight Obesity System Map and tested it against longitudinal data from the 1958 National Child Development Study. The framework divided study participants according to two categories of risk: 'conditioning factors' (past/fixed events and conditions) and 'intervention factors' (present and modifiable). At the age of 11 years, conditioning factors were 'low/high birth weight' and 'absence of breastfeeding', and intervention factors were 'low childhood activity level' and 'having at least one obese parent'. From a composite score of all four variables, study participants were assigned to one of the four risk groups: low risk, past 'conditioning' risk only, present 'intervention' risk only and high combined risk. ORs and relative risks for the development of future overweight/obesity at ages 23, 33 and 42 years were calculated for each risk group. Those identified in the highest risk category at the age of 11 were around twice as likely to become overweight (body mass index (BMI)≥25 kg/m(2)) by the age of 23 years, and obese (BMI≥30 kg/m(2)) by ages 33 and 42 years, in comparison to their low-risk peers (total sample, N=11 752). Increased prevalence of future obesity was also observed for high-risk children who were not already overweight at the age of 11 (filtered sample, N=9549). This framework identifies a greater proportion of the population that is at risk for future obesity than does childhood weight assessment alone.
    Journal of epidemiology and community health 09/2013; 67(12). DOI:10.1136/jech-2012-201978 · 3.50 Impact Factor
Show more