The Effects of Reverse Causality and Selective Attrition on the Relationship Between Body Mass Index and Mortality in Postmenopausal Women

Published on Oct 1, 2019in American Journal of Epidemiology4.473
· DOI :10.1093/aje/kwz160
Hailey R. Banack11
Estimated H-index: 11
Jennifer W. Bea16
Estimated H-index: 16
+ 10 AuthorsJean Wactawski-Wende67
Estimated H-index: 67
: Concerns about reverse causality and selection bias complicate the interpretation of studies of body mass index (BMI, calculated as weight (kg)/height (m)2) and mortality in older adults. The objective of this study was to investigate methodological explanations for the apparent attenuation of obesity-related risks in older adults. We used data from 68,132 participants in the Women's Health Initiative (WHI) clinical trial for this analysis. All of the participants were postmenopausal women aged 50-79 years at baseline (1993-1998). To examine reverse causality and selective attrition, we compared rate ratios from inverse probability of treatment- and censoring-weighted Poisson marginal structural models with results from an unweighted adjusted Poisson regression model. The estimated mortality rate ratios and 95% confidence intervals for BMIs of 30.0-34.9, 35.0-39.9 and ≥40.0 were 0.86 (95% confidence interval (CI): 0.77, 0.96), 0.85 (95% CI: 0.72, 0.99), and 0.88 (95% CI: 0.72, 1.07), respectively, in the unweighted model. The corresponding mortality rate ratios were 0.96 (95% CI: 0.86, 1.07), 1.12 (95% CI: 0.97, 1.29), and 1.31 95% CI: (1.08, 1.57), respectively, in the marginal structural model. Results from the inverse probability of treatment- and censoring-weighted marginal structural model were attenuated in low BMI categories and increased in high BMI categories. The results demonstrate the importance of accounting for reverse causality and selective attrition in studies of older adults.
  • References (63)
  • Citations (0)
#1Mohammad Ali Mansournia (Tehran University of Medical Sciences)H-Index: 28
#2Ashley I. Naimi (University of Pittsburgh)H-Index: 17
Last. Sander Greenland (UCLA: University of California, Los Angeles)H-Index: 106
view all 3 authors...
: There are now many published applications of causal (structural) models for estimating effects of time-varying exposures in the presence of confounding by earlier exposures and confounders affected by earlier exposures. Results from these models can be highly sensitive to inclusion of lagged and baseline exposure terms for different visits. This sensitivity is often overlooked in practice; moreover, results from these models are not directly comparable to results from conventional time-depende...
2 CitationsSource
#1Maryam Shakiba (University of Gilan)H-Index: 9
#2Mohammad Ali Mansournia (Tehran University of Medical Sciences)H-Index: 28
Last. Jay S. Kaufman (McGill University)H-Index: 64
view all 3 authors...
OBJECTIVE: This study quantified the obesity-stroke relationship by appropriately adjusting for time-varying confounders using G-estimation. METHODS: A total of 13,975 participants in the Atherosclerosis Risk in Communities (ARIC) study were included. General obesity (GOB) was defined as BMI ≥ 30 kg/m2 ; abdominal obesity (AOB) was defined as waist circumference ≥ 102 cm in men and ≥ 88 cm in women and waist to hip ratio ≥ 0.9 in men and ≥ 0.85 in women. The effects of obesity on stroke were est...
1 CitationsSource
#1Hailey R. Banack (UB: University at Buffalo)H-Index: 11
#2Andrew Stokes (BU: Boston University)H-Index: 23
Last. Jean Wactawski-Wende (UB: University at Buffalo)H-Index: 67
view all 17 authors...
Background:There is widespread concern about the use of body mass index (BMI) to define obesity status in postmenopausal women because it may not accurately represent an individual’s true obesity status. The objective of the present study is to examine and adjust for exposure misclassification bias
3 CitationsSource
#1Marte Karoline Råberg Kjøllesdal (University of Oslo)H-Index: 8
#2George Davey-SmithH-Index: 194
Last. Øyvind Næss (University of Oslo)H-Index: 24
view all 6 authors...
The article aims to describe the association between midlife body mass index (BMI) and cardiovascular disease (CVD)- and all-cause mortality, and to use early adulthood BMI as an instrumental variable for midlife BMI, in order to obtain an estimate less distorted by midlife confounders and reverse causality. Data from Norwegian health surveys (1974–2003) (midlife BMI, smoking, blood pressure, total cholesterol, heart rate), Military Conscription Records, National Tuberculosis Screenings (early a...
3 CitationsSource
#1Elizabeth Rose Mayeda (UCLA: University of California, Los Angeles)H-Index: 18
#2Hailey R. Banack (UB: University at Buffalo)H-Index: 11
Last. M. Maria Glymour (UCSF: University of California, San Francisco)H-Index: 43
view all 7 authors...
Background:In middle age, stroke incidence is higher among black than white Americans. For unknown reasons, this inequality decreases and reverses with age. We conducted simulations to evaluate whether selective survival could account for observed age patterning of black–white stroke inequalities.Me
4 CitationsSource
2 CitationsSource
#1Hailey R. Banack (McGill University)H-Index: 11
#2Sam Harper (McGill University)H-Index: 33
Last. Jay S. Kaufman (McGill University)H-Index: 64
view all 3 authors...
Abstract Background In cardiovascular research, pre-hospital mortality represents an important potential source of selection bias. Inverse probability of censoring weights are a method to account for this source of bias. The objective of this article is to examine and correct for the influence of selection bias due to pre-hospital mortality on the relationship between cardiovascular risk factors and all-cause mortality after an acute cardiac event. Methods The relationship between the number of ...
3 CitationsSource
#1Maryam Shakiba (University of Gilan)H-Index: 9
#2Mohammad Ali Mansournia (Tehran University of Medical Sciences)H-Index: 28
Last. Jay S. Kaufman (McGill University)H-Index: 64
view all 6 authors...
: In longitudinal studies, standard analysis may yield biased estimates of exposure effect in the presence of time-varying confounders that are also intermediate variables. We aimed to quantify the relationship between obesity and coronary heart disease (CHD) by appropriately adjusting for time-varying confounders. This study was performed in a subset of participants from the Atherosclerosis Risk in Communities (ARIC) Study (1987-2010), a US study designed to investigate risk factors for atheros...
8 CitationsSource
#1David Carslake (UoB: University of Bristol)H-Index: 10
#2George Davey-Smith (UoB: University of Bristol)H-Index: 194
Last. Pål Romundstad (NTNU: Norwegian University of Science and Technology)H-Index: 54
view all 6 authors...
Background: The observational association between mortality and body mass index (BMI) is U-shaped, leading to highly publicized suggestions that moderate overweight is beneficial to health. However, it is unclear whether elevated mortality is caused by low BMI or if the association is confounded, for example by concurrent ill health. Methods: Using HUNT, a Norwegian prospective study, 32 452 mother-offspring and 27 747 father-offspring pairs were followed up to 2009. Conventional hazard ratios f...
6 CitationsSource
#1Zhao Chen (UA: University of Arizona)H-Index: 33
#2Yann C. Klimentidis (UA: University of Arizona)H-Index: 17
Last. Cynthia A. Thomson (UA: University of Arizona)H-Index: 56
view all 7 authors...
Objectives To determine whether the relationship between anthropometric measurements of obesity and mortality varies according to age, race, and ethnicity in older women. Design Prospective cohort study of multiethnic postmenopausal women. Setting Women's Health Initiative (WHI) observational study and clinical trials in 40 clinics. Participants Postmenopausal women aged 50–79 participating in WHI (N = 161,808). Measurements Baseline height, weight, and waist circumference (WC) were measured, an...
12 CitationsSource
Cited By0
#1Lucile Migault (University of Bordeaux)H-Index: 4
#2Ronan Garlantézec (University of Rennes)H-Index: 2
Last. Ghislaine Bouvier (University of Bordeaux)H-Index: 13
view all 14 authors...
Background Data on the effects of extremely low frequency electromagnetic fields (ELF-EMF) on pregnancy outcomes are inconclusive. Objective To study the relation between maternal cumulative exposure to ELF-EMF during pregnancy and the risk of prematurity or small for gestational age (SGA) in a pooled analysis of two French birth cohorts. Methods Elfe and Epipage2 are both population-based birth cohorts initiated in 2011 and included 18 329 and 8400 births, respectively. Health data and househol...
1 CitationsSource