Influence of confounding factors in determining the risk of chronic lung disease among the elderly: Application of the mantel-haenszel test

Abstract

In epidemiological studies, the cause-and-effect relationship is commonly measured using the odds ratio (OR), which reflects the odds of an outcome occurring versus not occurring. However, the crude odds ratio is specific to individual strata and does not account for the influence of confounding variables. The Mantel-Haenszel (MH) test, an extension of the chi-square test, estimates the adjusted odds ratio by accounting for such confounders. This study attempts to assess the influence of confounding factors on the risk of chronic lung disease (CLD) among the elderly, using data from the Study on Global AGEing and Adult Health (SAGE), India, Wave 3. Data were collected from individuals aged 50 years and above who reported being diagnosed with hypertension (HT), diabetes (DI), and asthma (AS), considered as possible risk factors for CLD (defined as “ever diagnosed with chronic lung disease”). Tobacco use (“ever smoked or used smokeless tobacco,” SMOKE) and alcohol consumption (“ever consumed alcohol,” AL) were treated as confounding variables. In addition to the chi-square test, which assesses associations and crude odds ratios, the MH test was applied to estimate adjusted odds ratios and evaluate the magnitude of confounding. Since smoking and alcohol use vary by age and sex, the analysis was stratified by the age group (AGE) and sex (SEX). Analysis revealed that.