PURPOSE: The response of bone to exercise in women with anorexia nervosa (AN) is unclear. We investigated the associations between bone mineral density (BMD) and exercise performed while ill and while recovered in women with a history of AN. METHODS: A cross-sectional study was conducted with 141 women with AN (85 ill; 56 recovered), aged 17-40 yr. BMD at the lumbar spine (LS), femoral neck (FN), and total body (TB) was measured by dual-energy x-ray absorptiometry. Life History Calendar and Minnesota Leisure Time Physical Activity interviews were used to collect lifetime illness and exercise histories (amount and bone loading type). Average hours per week of each of moderate (MOD) and high (HI) bone loading exercise were determined for three illness phases: "before ill," "while ill," and "while recovered." Participants were categorized into four exercise groups for each phase: MOD-ONLY, HI-ONLY, BOTH, and NEITHER (reference group). Weight-adjusted BMD z-scores were compared in the exercise groups by multivariable linear regression, adjusting for illness duration and severity, and exercise during the other illness phases. RESULTS: In ill participants, MOD-ONLY "while ill" had lower BMD at LS (β = -0.69, 95% confidence interval (CI) = -1.02 to -0.05) and TB (β = -0.73, 95% CI = -1.31 to -0.15) than the NEITHER group. In recovered participants, HI-ONLY "while recovered" had higher BMD at FN (β = 0.95, 95% CI = 0.15-1.75) and TB (β = 0.79, 95% CI = 0.07-1.51) than the NEITHER group. CONCLUSIONS: The effect of exercise on bone in AN patients is dependent on both the type of mechanical loading and the phase of illness during which it was performed. Excessive moderate loading exercise while ill may put patients at higher risk of low bone mass, but high bone loading activities may provoke bone accrual during recovery.
PURPOSE: The response of bone to exercise in women with anorexia nervosa (AN) is unclear. We investigated the associations between bone mineral density (BMD) and exercise performed while ill and while recovered in women with a history of AN. METHODS: A cross-sectional study was conducted with 141 women with AN (85 ill; 56 recovered), aged 17-40 yr. BMD at the lumbar spine (LS), femoral neck (FN), and total body (TB) was measured by dual-energy x-ray absorptiometry. Life History Calendar and Minnesota Leisure Time Physical Activity interviews were used to collect lifetime illness and exercise histories (amount and bone loading type). Average hours per week of each of moderate (MOD) and high (HI) bone loading exercise were determined for three illness phases: "before ill," "while ill," and "while recovered." Participants were categorized into four exercise groups for each phase: MOD-ONLY, HI-ONLY, BOTH, and NEITHER (reference group). Weight-adjusted BMD z-scores were compared in the exercise groups by multivariable linear regression, adjusting for illness duration and severity, and exercise during the other illness phases. RESULTS: In ill participants, MOD-ONLY "while ill" had lower BMD at LS (β = -0.69, 95% confidence interval (CI) = -1.02 to -0.05) and TB (β = -0.73, 95% CI = -1.31 to -0.15) than the NEITHER group. In recovered participants, HI-ONLY "while recovered" had higher BMD at FN (β = 0.95, 95% CI = 0.15-1.75) and TB (β = 0.79, 95% CI = 0.07-1.51) than the NEITHER group. CONCLUSIONS: The effect of exercise on bone in AN patients is dependent on both the type of mechanical loading and the phase of illness during which it was performed. Excessive moderate loading exercise while ill may put patients at higher risk of low bone mass, but high bone loading activities may provoke bone accrual during recovery.
Authors: Melanie Schorr; Anne Drabkin; Micol S Rothman; Erinne Meenaghan; Gillian T Lashen; Margherita Mascolo; Ashlie Watters; Tara M Holmes; Kate Santoso; Elaine W Yu; Madhusmita Misra; Kamryn T Eddy; Anne Klibanski; Philip Mehler; Karen K Miller Journal: Clin Endocrinol (Oxf) Date: 2019-04-01 Impact factor: 3.478
Authors: Andrea Trombetti; Laura Richert; François R Herrmann; Thierry Chevalley; Jean-Daniel Graf; René Rizzoli Journal: Int J Endocrinol Date: 2013-03-24 Impact factor: 3.257