I Saito1, H Iso, Y Kokubo, M Inoue, S Tsugane. 1. Department of Public Health, Social Medicine and Medical Informatics, Ehime University Graduate School of Medicine, Toon, Japan. saitoi@m.ehime-u.ac.jp
Abstract
OBJECTIVE: The Japan Public Health Center-based prospective Study examined the association of body mass index (BMI) and weight change with incident stroke in Japanese individuals, for whom BMI levels are generally low. METHODS: We used initial data from 1990 to 1994 and 5-year follow-up surveys from 1995 to 1999. We calculated weight change over a 5-year period for 32,847 men and 38,875 women, aged 45-74 years, with no history of cardiovascular disease or cancer. Subjects were followed from the 1995-1999 survey to the end of 2005, and hazard ratios of self-reported BMI levels and weight change for incident stroke were estimated using Cox's proportional hazard models adjusted for potential confounders. RESULTS: During the follow-up period (median 7.9 years) there were 2019 incident strokes, including subtypes. Multivariable-adjusted hazard ratios for all stroke events by BMI levels of 27.0-29.9 and ≥ 30 kg m(-2) versus 23.0-24.9 kg m(-2) were 1.09 (95% confidence interval 0.88, 1.36) and 1.25 (0.86, 1.84) in men (P for trend=0.22), and 1.29 (1.01, 1.65) and 2.16 (1.60, 2.93) in women (P for trend <0.001), respectively. A weight change of ≥ 10% in the previous 5 years was associated with total strokes and ischemic strokes in women. CONCLUSION: Higher BMI levels and a weight gain of ≥ 10% over 5 years were associated with an increased risk of stroke in women, whereas this association was weak in men.
OBJECTIVE: The Japan Public Health Center-based prospective Study examined the association of body mass index (BMI) and weight change with incident stroke in Japanese individuals, for whom BMI levels are generally low. METHODS: We used initial data from 1990 to 1994 and 5-year follow-up surveys from 1995 to 1999. We calculated weight change over a 5-year period for 32,847 men and 38,875 women, aged 45-74 years, with no history of cardiovascular disease or cancer. Subjects were followed from the 1995-1999 survey to the end of 2005, and hazard ratios of self-reported BMI levels and weight change for incident stroke were estimated using Cox's proportional hazard models adjusted for potential confounders. RESULTS: During the follow-up period (median 7.9 years) there were 2019 incident strokes, including subtypes. Multivariable-adjusted hazard ratios for all stroke events by BMI levels of 27.0-29.9 and ≥ 30 kg m(-2) versus 23.0-24.9 kg m(-2) were 1.09 (95% confidence interval 0.88, 1.36) and 1.25 (0.86, 1.84) in men (P for trend=0.22), and 1.29 (1.01, 1.65) and 2.16 (1.60, 2.93) in women (P for trend <0.001), respectively. A weight change of ≥ 10% in the previous 5 years was associated with total strokes and ischemic strokes in women. CONCLUSION: Higher BMI levels and a weight gain of ≥ 10% over 5 years were associated with an increased risk of stroke in women, whereas this association was weak in men.
Authors: Mary E Kroll; Jane Green; Valerie Beral; Cathie L M Sudlow; Anna Brown; Oksana Kirichek; Alison Price; TienYu Owen Yang; Gillian K Reeves Journal: Neurology Date: 2016-09-07 Impact factor: 9.910
Authors: Zhengming Chen; Andri Iona; Sarah Parish; Yiping Chen; Yu Guo; Fiona Bragg; Ling Yang; Zheng Bian; Michael V Holmes; Sarah Lewington; Ben Lacey; Ruqin Gao; Fang Liu; Zengzhi Zhang; Junshi Chen; Robin G Walters; Rory Collins; Robert Clarke; Richard Peto; Liming Li Journal: Lancet Glob Health Date: 2018-06 Impact factor: 26.763