Original article
Does selection bias explain the obesity paradox among individuals with cardiovascular disease?

https://doi.org/10.1016/j.annepidem.2015.02.008Get rights and content

Highlights

  • • This article highlights the message that the obesity paradox may be simply a statistical artifact, a product of selection bias (also known as collider stratification bias) rather than a true effect of clinical relevance

  • • The article adds to the current body of literature on the obesity paradox by providing numeric evidence from sensitivity analyses supporting the hypothesis that selection bias provides a reasonable explanation for the obesity paradox

  • • This article has important implications for clinicians treating obese individuals with chronic disease as it underscores the fact that they should not change their current practice based on articles claiming to have found evidence of a true obesity paradox

Abstract

Objectives

The objectives of this article are to demonstrate that the obesity paradox may be explained by collider stratification bias and to estimate the biasing effects of unmeasured common causes of cardiovascular disease (CVD) and mortality on the observed obesity-mortality relationship.

Methods

We use directed acyclic graphs, regression modeling, and sensitivity analyses to explore whether the observed protective effect of obesity among individuals with CVD can be plausibly attributed to selection bias. Data from the third National Health and Examination Survey was used for the analyses.

Results

The adjusted total effect of obesity on mortality was a risk difference (RD) of 0.03 (95% confidence interval [CI]: 0.02, 0.05). However, the controlled direct effect of obesity on mortality among individuals without CVD was RD = 0.03 (95% CI: 0.01, 0.05) and RD = −0.12 (95% CI: −0.20, −0.04) among individuals with CVD. The adjusted total effect estimate demonstrates an increased number of deaths among obese individuals relative to nonobese counterparts, whereas the controlled direct effect shows a paradoxical decrease in morality among obese individuals with CVD.

Conclusions

Sensitivity analysis demonstrates unmeasured confounding of the mediator-outcome relationship provides a sufficient explanation for the observed protective effect of obesity on mortality among individuals with CVD.

Section snippets

Graphical representation of the obesity paradox

The directed acyclic graph presented in Figure 1 depicts the hypothesized relations between obesity, CVD, and all-cause mortality. CVD is known as an intermediate (or mediator) variable, as it is on the causal path from obesity to mortality [25], [26]. Let C represent a vector of measured covariates that confound the relationship between obesity and mortality (e.g., age, gender). For the purpose of this article, we will use cardiorespiratory fitness (CRF) as an example of a common cause of CVD

Data source

The obesity-mortality relationship was analyzed using data from the third National Health and Nutrition Examination Survey (NHANES III; 1988–1994), a nationally representative survey of noninstitutionalized civilians in the United States [39]. The study population is recruited through a stratified multistage probability sampling procedure, designed to represent the entire U.S. population [39]. NHANES III participants completed a standardized in-home interview and questionnaire as well as a

Sensitivity analysis

To explore the possibility that the protective effect of obesity on mortality among individuals with CVD is because of unmeasured confounding of the CVD-mortality relationship by CRF (Fig. 1), we conducted sensitivity analyses for the CDE [18], [47]. If Yobese,cvd is independent of Obese|[C, CRF] and Yobese,cvd is independent of CVD|[Obese, C, CRF] for all levels of obesity and CVD, the magnitude of bias in the CDE is given by:Bias(CDEobese,obese*(CVD))=δγwhere γ is equal to E[mortality|obese,

Acknowledgment

The authors acknowledge Rebecca Anthopolos for her help in creating the graphics for this article.

Hailey Banack was supported by a doctoral research award from the Fonds de la Recherche en Sante du Quebec and a CIHR Institute of Circulatory and Respiratory Health Skills Development Award. Jay Kaufman was supported by the Canada Research Chair program.

References (78)

  • J.B. Kampert et al.

    Physical activity, physical fitness, and all-cause and cancer mortality: a prospective study of men and women

    Ann Epidemiol

    (1996)
  • P. Cuijpers et al.

    Excess mortality in depression: a meta-analysis of community studies

    J Affective Disord

    (2002)
  • T.B. Harris et al.

    Associations of elevated interleukin-6 and C-reactive protein levels with mortality in the elderly

    Am J Med

    (1999)
  • K.M. Flegal et al.

    Comparisons of percentage body fat, body mass index, waist circumference, and waist-stature ratio in adults

    Am J Clin Nutr

    (2009)
  • A. De Schutter et al.

    The impact of obesity on risk factors and prevalence and prognosis of coronary heart disease—the obesity paradox

    Prog Cardiovasc Dis

    (2014)
  • C.J. Lavie et al.

    Body composition and coronary heart disease mortality—an obesity or a lean paradox?

    Mayo Clin Proc

    (2011)
  • A. De Schutter et al.

    Body composition and mortality in a large cohort with preserved ejection fraction: untangling the obesity paradox

    Mayo Clin Proc

    (2014)
  • Y. Okura et al.

    Agreement between self-report questionnaires and medical record data was substantial for diabetes, hypertension, myocardial infarction and stroke but not for heart failure

    J Clin Epidemiol

    (2004)
  • K.F. Adams et al.

    Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old

    N Engl J Med

    (2006)
  • E.E. Calle et al.

    Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults

    N Engl J Med

    (2003)
  • K.M. Flegal et al.

    Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis

    JAMA

    (2013)
  • P.A. McAuley et al.

    Obesity paradoxes

    J Sports Sci

    (2011)
  • H.R. Banack et al.

    Reversing the direction of associations among survivors: the “obesity paradox” explained

    Epidemiology

    (2013)
  • M.R. Carnethon et al.

    Association of weight status with mortality in adults with incident diabetes

    JAMA

    (2012)
  • K. Kalantar-Zadeh et al.

    Mortality prediction by surrogates of body composition: an examination of the obesity paradox in hemodialysis patients using composite ranking score analysis

    Am J Epidemiol

    (2012)
  • S. Schlesinger et al.

    Postdiagnosis body mass index and risk of mortality in colorectal cancer survivors: a prospective study and meta-analysis

    Cancer Causes Control

    (2014)
  • A. Amadou et al.

    Overweight, obesity and risk of premenopausal breast cancer according to ethnicity: a systematic review and dose-response meta-analysis

    Obes Rev

    (2013)
  • M.M. Glymour et al.

    Commentary: selection bias as an explanation for the obesity paradox: just because it's possible doesn't mean it's plausible

    Epidemiology

    (2014)
  • M. Lajous et al.

    Body mass index, diabetes, and mortality in French women: explaining away a “paradox”

    Epidemiology

    (2014)
  • Lajous M, Banack HR, Kaufman JS, Hernán MA. Should patients with chronic disease be told to gain weight? The obesity...
  • M.A. Hernán et al.

    A structural approach to selection bias

    Epidemiology

    (2004)
  • T.J. VanderWeele

    Bias formulas for sensitivity analysis for direct and indirect effects

    Epidemiology

    (2010)
  • S.H. Preston et al.

    Obesity paradox: conditioning on disease enhances biases in estimating the mortality risks of obesity

    Epidemiology

    (2014)
  • H.R. Banack et al.

    Perhaps the correct answer is: (D) all of the above

    Epidemiology

    (2014)
  • D.W. Flanders et al.

    A nearly unavoidable mechanism for collider bias in the obesity-end-stage-renal-disease-mortality and similar studies

    Epidemiology

    (2014)
  • S. Hernández-Díaz et al.

    The birth weight “paradox” uncovered?

    Am J Epidemiol

    (2006)
  • O. Basso et al.

    Birth weight and mortality: causality or confounding?

    Am J Epidemiol

    (2006)
  • N.J. Pagidipati et al.

    Estimating deaths from cardiovascular disease: a review of global methodologies of mortality measurement

    Circulation

    (2013)
  • S. Kodama et al.

    Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis

    JAMA

    (2009)
  • Cited by (0)

    View full text