A 71-year-old man comes to the office due to slowly progressive shortness of breath for the past year. He enjoys working in his garden, but shortness of breath has forced him to reduce the time he spends there. He also has a mild morning cough productive of whitish sputum. He estimates that his physical activity level is <30 min/day. The patient has a history of hypertension. He is a retired middle school teacher. He smoked 2 packs of cigarettes daily for 20 years and quit 15 years ago. Blood pressure is 127/85 mm Hg, pulse is 78/min, respirations are 14/min, and pulse oximetry is 93% on room air. BMI is 24 kg/m². Occasional end-expiratory wheezes are present on lung examination, and the heart examination demonstrates a regular rate and rhythm. Chest x-ray shows hyperinflation, and pulmonary function test results are consistent with an obstructive defect.
Based on the results of the study, which of the following is most likely to occur if the patient increases his physical activity level to >60 min/day?

Question

Quizplus · Accepted Answer

Explanation: The odds ratio (OR) is a measure of association between exposure (eg, risk factor) and outcome (eg, disease), with OR >1 meaning that exposure is associated with higher odds of outcome, OR <1 meaning that exposure is associated with lower odds of outcome, and OR = 1 meaning that exposure has no effect on odds of outcome. A 95% confidence interval (CI) that does not contain the null value (in this case, OR = 1) is often used as a proxy for statistical significance when the p-value is not provided. In this example, univariate analyses revealed that all risk factors apart from anxiety were statistically associated with high-level (ie, ≥60 min/day) physical activity (PA) given CIs excluding the null value (for anxiety, the association was nonstatistically significant given 95% CI: 0.13, 1.05). However, univariate analysis does not take into account confounders, additional independent variables that may also be associated with exposure and outcome and that may mask the real association between exposure and outcome if not adjusted. Multiple regression is one of the techniques used during the data analysis phase of a study to adjust for confounders; it takes into account the effect from additional independent variables included in the analysis and estimates "adjusted" ORs. An adjusted OR provides information about the strength of the association between exposure and outcome once all other potential confounders that might mask the real association have been controlled for in the regression model. Based on the abstract, after adjusting for all independent variables considered, high-level PA was significantly associated only with depression based on HADS-D (OR = 0.48, 95% CI: 0.24, 0.98) and dyspnea based on the MMRC scale (OR = 0.39, 95% CI: 0.17, 0.93) (Choices A, B, D, and E). This means that, if the patient increases his PA level to ≥60 min/day, he will have a statistically significant decrease (because OR <1) in dyspnea based on his MMRC score and in depression based on his HADS-D score. Educational objective: Confounding bias affects the statistical analysis of study results. Confounders may be controlled during the data analysis phase with multiple regression. Factors statistically associated with the outcome in univariate analysis may not necessarily be significant in the multivariable analysis once all other factors have been adjusted for in the model.

A 71-Year-Old Man Comes to the Office Due to Slowly