The Effect of Changes in Cardiorespiratory Fitness and Weight on Obstructive Sleep Apnea Severity in Overweight Adults with Type 2 Diabetes

• Published in: Sleep (New York, N.Y.) N.Y.), 2016-02, Vol.39 (2), p.317-325
• Main Authors: Kline, Christopher E, Reboussin, David M, Foster, Gary D, Rice, Thomas B, Strotmeyer, Elsa S, Jakicic, John M, Millman, Richard P, Pi-Sunyer, F Xavier, Newman, Anne B, Wadden, Thomas A, Zammit, Gary, Kuna, Samuel T
• Format: Article
• Language: English
• Subjects: Aged; Body Mass Index; Cardiovascular Physiological Phenomena; Diabetes Mellitus, Type 2 - complications; Diabetes Mellitus, Type 2 - physiopathology; Female; Health Status; Humans; Life Style; Male; Middle Aged; Obesity - complications; Obesity - physiopathology; Overweight - complications; Overweight - physiopathology; Patient Education as Topic; Physical Fitness - physiology; Respiratory Physiological Phenomena; Sleep Apnea, Obstructive - complications; Sleep Apnea, Obstructive - diagnosis; Sleep Apnea, Obstructive - physiopathology; Sleep Disordered Breathing; Treatment Outcome; Weight Loss
• ISSN: 0161-8105; 1550-9109
• DOI: 10.5665/sleep.5436

To examine the effect of changes in cardiorespiratory fitness on obstructive sleep apnea (OSA) severity prior to and following adjustment for changes in weight over the course of a 4-y weight loss intervention. As secondary analyses of a randomized controlled trial, 263 overweight/obese adults with type 2 diabetes and OSA participated in an intensive lifestyle intervention or education control condition. Measures of OSA severity, cardiorespiratory fitness, and body weight were obtained at baseline, year 1, and year 4. Change in the apnea-hypopnea index (AHI) served as the primary outcome. The percentage change in fitness (submaximal metabolic equivalents [METs]) and change in weight (kg) were the primary independent variables. Primary analyses collapsed intervention conditions with statistical adjustment for treatment group and baseline METs, weight, and AHI among other relevant covariates. At baseline, greater METs were associated with lower AHI (B [SE] = -1.48 [0.71], P = 0.038), but this relationship no longer existed (B [SE] = -0.24 [0.73], P = 0.75) after adjustment for weight (B [SE] = 0.31 [0.07], P < 0.0001). Fitness significantly increased at year 1 (+16.53 ± 28.71% relative to baseline), but returned to near-baseline levels by year 4 (+1.81 ± 24.48%). In mixed-model analyses of AHI change over time without consideration of weight change, increased fitness at year 1 (B [SE] = -0.15 [0.04], P < 0.0001), but not at year 4 (B [SE] = 0.04 [0.05], P = 0.48), was associated with AHI reduction. However, with weight change in the model, greater weight loss was associated with AHI reduction at years 1 and 4 (B [SE] = 0.81 [0.16] and 0.60 [0.16], both P < 0.0001), rendering the association between fitness and AHI change at year 1 nonsignificant (B [SE] = -0.04 [0.04], P = 0.31). Among overweight/obese adults with type 2 diabetes, fitness change did not influence OSA severity change when weight change was taken into account. ClinicalTrials.gov identification number NCT00194259.