Hostname: page-component-78c5997874-xbtfd Total loading time: 0 Render date: 2024-11-13T02:57:02.618Z Has data issue: false hasContentIssue false
  • English
  • Français

A comparison of confounding adjustment methods with an application to early life determinants of childhood obesity

Published online by Cambridge University Press:  29 August 2014

L. Li ,
K. Kleinman  and
M. W. Gillman
L. Li*
Affiliation:
Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, USA
K. Kleinman
Affiliation:
Department of Population Medicine, Obesity Prevention Program, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, USA
M. W. Gillman
Affiliation:
Department of Population Medicine, Obesity Prevention Program, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, USA
*
*Address for correspondence: L. Li, Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, 133 Brookline Avenue, 6th floor, Boston, MA 02215, USA.(Email Lingling_li@post.harvard.edu)
Get access Rights & Permissions [Opens in a new window]

Abstract

We implemented six confounding adjustment methods: (1) covariate-adjusted regression, (2) propensity score (PS) regression, (3) PS stratification, (4) PS matching with two calipers, (5) inverse probability weighting and (6) doubly robust estimation to examine the associations between the body mass index (BMI) z-score at 3 years and two separate dichotomous exposure measures: exclusive breastfeeding v. formula only (n=437) and cesarean section v. vaginal delivery (n=1236). Data were drawn from a prospective pre-birth cohort study, Project Viva. The goal is to demonstrate the necessity and usefulness, and approaches for multiple confounding adjustment methods to analyze observational data. Unadjusted (univariate) and covariate-adjusted linear regression associations of breastfeeding with BMI z-score were −0.33 (95% CI −0.53, −0.13) and −0.24 (−0.46, −0.02), respectively. The other approaches resulted in smaller n (204–276) because of poor overlap of covariates, but CIs were of similar width except for inverse probability weighting (75% wider) and PS matching with a wider caliper (76% wider). Point estimates ranged widely, however, from −0.01 to −0.38. For cesarean section, because of better covariate overlap, the covariate-adjusted regression estimate (0.20) was remarkably robust to all adjustment methods, and the widths of the 95% CIs differed less than in the breastfeeding example. Choice of covariate adjustment method can matter. Lack of overlap in covariate structure between exposed and unexposed participants in observational studies can lead to erroneous covariate-adjusted estimates and confidence intervals. We recommend inspecting covariate overlap and using multiple confounding adjustment methods. Similar results bring reassurance. Contradictory results suggest issues with either the data or the analytic method.

Keywords

breastfeedingcesarean sectionconfounding adjustmentobesitypropensity score
Type
Original Article
Information
Journal of Developmental Origins of Health and Disease , Volume 5 , Supplement 6 , December 2014 , pp. 435 - 447
Copyright
© Cambridge University Press and the International Society for Developmental Origins of Health and Disease 2014 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1. Imbens, GW, Angrist, JD. Identification and estimation of local average treatment effects. Econometrica. 1994; 62, 467–475.CrossRefGoogle Scholar
2. Lash, TL, Fox, MP, Fink, AK. Applying Quantitative Bias Analysis to Epidemiologic Data, 2009. Springer New York: New York, NY.Google Scholar
3. Martens, EP, Pestman, WR, de Boer, A, Belitser, SV, Klungel, OH. Instrumental variables: application and limitations. Epidemiology. 2006; 17, 260–267.CrossRefGoogle ScholarPubMed
4. Rubin, DB Estimating causal effects from large data sets using propensity scores. Ann Intern Med. 1997; 127(Pt 2), 757–763.CrossRefGoogle ScholarPubMed
5. Rosenbaum, PR, Rubin, DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983; 70, 41–55.CrossRefGoogle Scholar
6. Robins, JM, Ritov, Y. Toward a curse of dimensionality appropriate (CODA) asymptotic theory for semi-parametric models. Stat Med. 1997; 16, 285–319.3.0.CO;2-#>CrossRefGoogle Scholar
7. Ho, DE, Imai, K, King, G, Stuart, EA. Matching as nonparametric preprocessing for reducing model dependence in parametric causal inference. Polit Anal. 2007; 15, 199–236.CrossRefGoogle Scholar
8. Glynn, RJ, Schneeweiss, S, Sturmer, T. Indications for propensity scores and review of their use in pharmacoepidemiology. Basic Clin Pharmacol Toxicol. 2006; 98, 253–259.CrossRefGoogle ScholarPubMed
9. Sturmer, T, Joshi, M, Glynn, RJ, et al. A review of the application of propensity score methods yielded increasing use, advantages in specific settings, but not substantially different estimates compared with conventional multivariable methods. J Clin Epidemiol. 2006; 59, 437–447.CrossRefGoogle Scholar
10. Austin, PC. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat. 2010; 10, 150–161.CrossRefGoogle Scholar
11. Austin, PC. The performance of different propensity-score methods for estimating relative risks. J Clin Epidemiol. 2008; 61, 537–545.CrossRefGoogle ScholarPubMed
12. Austin, PC, Mamdani, MM, Stukel, TA, Anderson, GM, Tu, JV. The use of the propensity score for estimating treatment effects: administrative versus clinical data. Stat Med. 2005; 24, 1563–1578.Google Scholar
13. Austin, PC. A critical appraisal of propensity-score matching in the medical literature between 1996 and 2003. Stat Med. 2008; 27, 2037–2049.CrossRefGoogle ScholarPubMed
14. Stuart, EA. Developing practical recommendations for the use of propensity scores: discussion of ‘A critical appraisal of propensity score matching in the medical literature between 1996 and 2003’ by Peter Austin, Statistics in Medicine. Stat Med. 2008; 27, 2062–2065, discussion 2066–2069.CrossRefGoogle Scholar
15. Casella, G, Berger, RL. Statistical Inference, (vol. 2) 2002. Duxbury: Pacific Grove, CA.Google Scholar
16. Kurth, T, Walker, AM, Glynn, RJ, et al. Results of multivariate logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect. Am J Epidemiol. 2005; 163, 262–270.CrossRefGoogle ScholarPubMed
17. Rosenbaum, PR, Rubin, DB. Reducing bias in observational studies using subclassification on the propensity score. J Am Stat Assoc. 1984; 79, 516–524.CrossRefGoogle Scholar
18. Hernan, MA, Brumback, B, Robins, JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology. 2000; 11, 561–570.CrossRefGoogle ScholarPubMed
19. Robins, JM, Hernan, MA, Brumback, B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000; 11, 550–560.CrossRefGoogle ScholarPubMed
20. Bang, H, Robins, JM. Doubly robust estimation in missing data and causal inference models. Biometrics. 2005; 61, 962–972.CrossRefGoogle ScholarPubMed
21. van Rossem, L, Taveras, EM, Gillman, MW, et al. Is the association of breastfeeding with child obesity explained by infant weight change? Int J Pediatr Obes. 2011; 6, e415–e422, Epub 17472010 Oct 17477128.CrossRefGoogle ScholarPubMed
22. Owen, CG, Martin, RM, Whincup, PH, et al. The effect of breastfeeding on mean body mass index throughout life: a quantitative review of published and unpublished observational evidence. Am J Clin Nutr. 2005; 82, 1298–1307.CrossRefGoogle ScholarPubMed
23. Owen, CG, Martin, RM, Whincup, PH, Smith, GD, Cook, DG. Effect of infant feeding on the risk of obesity across the life course: a quantitative review of published evidence. Pediatrics. 2005; 115, 1367–1377.CrossRefGoogle ScholarPubMed
24. Gillman, MW. Commentary: breastfeeding and obesity – the 2011 Scorecard. Int J Epidemiol. 2011; 40, 681–684.CrossRefGoogle ScholarPubMed
25. Huh, SY, Rifas-Shiman, SL, Zera, CA, et al. Delivery by caesarean section and risk of obesity in preschool age children: a prospective cohort study. Arch Dis Child. 2012; 97, 610–616.CrossRefGoogle ScholarPubMed
26. Li, HT, Zhou, YB, Liu, JM. The impact of cesarean section on offspring overweight and obesity: a systematic review and meta-analysis. Int J Obes. 2013; 37(7), 893–899.CrossRefGoogle ScholarPubMed
27. Kramer, MS, Chalmers, B, Hodnett, ED, et al. Promotion of Breastfeeding Intervention trial (PROBIT): a randomized trial in the Republic of Belarus. JAMA. 2001; 285, 413–420.Google Scholar
28. Kramer, MS, Moodie, EE, Dahhou, M, Platt, RW. Breastfeeding and infant size: evidence of reverse causality. Am J Epidemiol. 2011; 173, 978–983.CrossRefGoogle ScholarPubMed
29. Lunceford, JK, Davidian, M. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat Med. 2004; 23, 2937–2960.CrossRefGoogle ScholarPubMed
30. Imbens, GW. Nonparametric estimation of average treatment effects under exogeneity: a review. Rev Econ Stat. 2004; 86, 4–29.CrossRefGoogle Scholar
31. Dehejia, RH, Wahba, S. Propensity score-matching methods for nonexperimental causal studies. Rev Econ Stat. 2002; 84, 151–161.CrossRefGoogle Scholar
32. Abadie, A, Imbens, GW. Large sample properties of matching estimators for average treatment effects. Econometrica. 2006; 74, 235–267.CrossRefGoogle Scholar
33. Hernan, MA, Cole, SR. Invited commentary: causal diagrams and measurement bias. Am J Epidemiol. 2009; 170, 959–962, discussion 963–954.CrossRefGoogle ScholarPubMed
34. Funk, MJ, Westreich, D, Davidian, M, Weisen, C. Introducing a SAS ® macro for doubly robust estimation. SAS Global Forum 2007, SAS, Inc., Orlando, Florida, 2007.Google Scholar
35. Gillman, MW, Rich-Edwards, JW, Rifas-Shiman, SL, et al. Maternal age and other predictors of newborn blood pressure. J Pediatr. 2004; 144, 240–245.CrossRefGoogle ScholarPubMed
36. Kuczmarski, RJ, Ogden, CL, Grummer-Strawn, LM, et al. CDC growth charts: United States. Advance data. 2000; 314, 1–27.Google Scholar
37. Sekhon, JS. Multivariate and propensity score matching software with automated balance optimization: the matching package for R. J Stat Softw. 2011; 42, 1–52.CrossRefGoogle Scholar
38. Rosenbaum, P. Observational Studies, 2002. Springer-Verlag: New York.CrossRefGoogle Scholar
39. Brumback, BA, Hernan, MA, SJPA, Haneuse, Robins, JM. Sensitivity analyses for unmeasured confounding assuming a marginal structural model for repeated measures. Stat Med. 2004; 23, 749–767.CrossRefGoogle ScholarPubMed
40. Li, L, Shen, CY, Wu, AC, Li, X. Propensity score-based sensitivity analysis method for uncontrolled confounding. Am J Epidemiol. 2011; 174, 345–353.Google Scholar
41. Robins, JM, Rotnitzky, A, Scharfstein, DO. Sensitivity analysis for selection bias and unmeasured confounding in missing data and causal inference models. In Statistical Models in Epidemiology: The Environment and Clinical Trials (eds. Halloran ME, Berry D), 1999; pp. 1–92. Springer-Verlag: New York.Google Scholar
42. Shen, CY, Li, X, Li, L, Were, MC. Sensitivity analysis for causal inference using inverse probability weighting. Biom J. 2011; 53, 822–837.CrossRefGoogle ScholarPubMed
Supplementary material: File

Li Supplementary Material

Supplementary Material

Download Li Supplementary Material(File)
File 20.2 KB