Obesity

Does waist circumference uncorrelated with BMI add valuable information? Gerard Ngueta,1 Elhadji A Laouan-Sidi,1 Michel Lucas1,2 ▸ Additional material is published online only. To view please visit the journal online (http://dx.doi.org/10.1136/jech2014-204005). 1

Population Health and Optimal Health Practices Research Unit, CHU de Québec Research Centre, Québec, Québec, Canada 2 Department of Social and Preventive Medicine, Faculty of Medicine, Laval University, Québec, Québec, Canada Correspondence to Dr Michel Lucas, Population Health and Optimal Health Practices Research Unit, CHU de Québec Research Centre, 2875 Laurier Blvd., Delta 2 Building, Office 600, Québec, Québec, Canada G1V 2M2; [email protected] Received 10 February 2014 Revised 13 May 2014 Accepted 21 May 2014 Published Online First 10 June 2014

ABSTRACT Background Estimation of relative contribution of Body Mass Index (BMI) and waist circumference (WC) on health outcomes requires a regression model that includes both obesity metrics. But, multicollinearity could yield biased estimates. Methods To address the multicollinearity issue between BMI and WC, we used the residual model approach. The standard WC (Y-axis) was regressed on the BMI (X-axis) to obtain residual WC. Data from two adult population surveys (Nunavik Inuit and James Bay Cree) were analysed to evaluate relative effect of BMI and WC on four cardiometabolic risk factors: insulin, triglycerides, systolic blood pressure and high-density lipoprotein levels. Results In multivariate models, standard WC and BMI were significantly associated with cardiometabolic outcomes. Residual WC was not linked with any outcomes. The BMI effect was weakened by including standard WC in the model, but its effect remained unchanged if residual WC was considered. Conclusions The strong correlation between standard WC and BMI does not allow assessment of their relative contributions to health in the same model without a risk of making erroneous estimations. By contrast with BMI, fat distribution (residual WC) does not add valuable information to a model that already contains overall adiposity (BMI) in Inuit and Cree.

INTRODUCTION

To cite: Ngueta G, LaouanSidi EA, Lucas M. J Epidemiol Community Health 2014;68:849–855.

The measures employed most frequently to identify excess weight in epidemiological surveys and clinical practice—Body Mass Index (BMI) and waist circumference (WC)—are part of a debate as to which of them best defines obesity. Developing anthropometric measures that could adequately foresee obesity is challenging.1 While some promote BMI as a valid measure of obesity,2 others favour WC as a better obesity metric, especially abdominal obesity.3 BMI and WC could contribute independently to several cardiovascular risk factors (such as hypertension, impairment of glucose and lipid metabolism),4–6 and are reported to be independent risk factors of diabetes,7 coronary heart disease8–10 and related mortality.11–14 BMI and WC are not independent of one another. Bouchard1 reported strong correlations between obesity-related anthropometrics, independently of sex and ethnicity. Després suggested that BMI and WC are not interchangeable at the individual level.15 Ideally, BMI and WC should be regressed in the same association model to estimate their relative contributions to health outcomes. Unfortunately, multicollinearity could yield biased regression estimates. Willett and Stampfer16 proposed the nutrient residual model to deal with

Ngueta G, et al. J Epidemiol Community Health 2014;68:849–855. doi:10.1136/jech-2014-204005

collinearity between nutrient and total caloric intake. The potential benefits of this methodology have not yet been explored with obesity-related anthropometrics. To evaluate the uncorrelated relative contributions of the two most commonly used obesity metrics, the first aim of the present work was to develop WC measures in which variations due to BMI were removed. We addressed this issue by regressing WC (Y-axis) on BMI (X-axis) to obtain residual WC. Then, the model included BMI and residual WC, which conceptually represented overall adiposity (BMI) and fat distribution (residual WC). This may help to isolate the direct and uncorrelated effects of both measures on different outcomes. To establish the usefulness of these two uncorrelated obesity measures, we accessed data from two population surveys to evaluate the association between BMI and WC on four cardiometabolic risk factors: insulin, triglycerides, systolic blood pressure (SBP) and high-density lipoprotein levels (HDL-C).

METHODS AND PROCEDURES Study populations The data analysed in this study were collected through the Nunavik Inuit Health Survey (2004) and the Multi-community Environment and Health Longitudinal Study (2005–2009) of the Eastern James Bay Cree. The designs of these studies have been described elsewhere.17 18 Among participants who were asked to undergo clinical examination, the Inuit survey included 914 Inuit adults (aged 18– 74 years), whereas, the James Bay Cree survey comprised 1001 adults (aged 18–89 years). After excluding pregnant women, and participants with missing information on BMI and WC, data from 810 Inuit and 833 Cree adults were available for analysis. Participation was voluntary and subject to written consent. Consent and assent forms were approved by the Comité d’éthique de la recherche de l’Université Laval and the Comité d’éthique de santé publique du Québec, and accepted by the research ethics boards of McMaster University and McGill University, as well as the research committee of the Cree Board of Health and Social Services of James Bay.

Data collection Face-to-face questionnaires were administered to participants to gather general sociodemographic data and information on lifestyle, such as tobacco and alcohol consumption. Clinical questionnaires documented medical history, anthropometric and physiological measures. Data were collected by trained nurses according to standard protocols, and included anthropometric measures (height, weight, 849

Obesity BMI) and sitting blood pressure. Mean SBP and diastolic blood pressure (DBP) were calculated from the last two measurements. A fasting blood sample was drawn from all participants. Within 3 h of sample collection, the tubes were labelled and refrigerated at 4°C prior to centrifugation. They were stored at −80°C until analysis at the Lipid Research Centre (CHUQ, Centre Hospitalier Universitaire de Québec). Triglyceride levels (mmol/ L) were quantified by enzymatic methods in a Vitros 950 Chemistry Station (Ortho-Clinical Diagnostics, Raritan, New Jersey, USA), employing the manufacturer’s reagents and calibrators. Samples were analysed by multilayer film dry-slide chemistry with colorimetric detection. Fasting plasma insulin ( pmol/L) was ascertained with the chemiluminescent detection immunoassay system (ADVIA Centaur CP Immunoassay system; Bayer HealthCare, Toronto, Ontario, Canada). Insulin levels ≥90 pmol/L defined hyperinsulinaemia, SBP≥130 mm Hg indicated high SBP, and blood triglyceride levels ≥1.7 mmol/L denoted hypertriglyceridaemia. Low HDL-C was defined as HDL-C

Does waist circumference uncorrelated with BMI add valuable information?

Estimation of relative contribution of Body Mass Index (BMI) and waist circumference (WC) on health outcomes requires a regression model that includes...
1MB Sizes 4 Downloads 3 Views