Menni Cristina, Migaud Marie, Kastenmüller Gabi, Pallister Tess, Zierer Jonas, Peters Annette, Mohney Robert P, Spector Tim D, Bagnardi Vincenzo, Gieger Christian, Moore Steve C, Valdes Ana M
Department of Twin Research & Genetic Epidemiology, King's College London, London, UK.
Mitchell Cancer Institute, University of South Alabama, Mobile, Alabama, USA.
Obesity (Silver Spring). 2017 Sep;25(9):1618-1624. doi: 10.1002/oby.21922. Epub 2017 Jul 31.
To investigate the association between long-term weight change and blood metabolites.
Change in BMI over 8.6 ± 3.79 years was assessed in 3,176 females from the TwinsUK cohort (age range: 18.3-79.6, baseline BMI: 25.11 ± 4.35) measured for 280 metabolites at follow-up. Statistically significant metabolites (adjusting for covariates) were included in a multivariable least absolute shrinkage and selection operator (LASSO) model. Findings were replicated in the Cooperative Health Research in the Region of Augsburg (KORA) study (n = 1,760; age range: 25-70, baseline BMI: 27.72 ± 4.53). The study examined whether the metabolites identified could prospectively predict weight change in KORA and in the Prostate, Lung, Colorectal, and Ovarian Cancer (PLCO) study (n = 471; age range: 55-74, baseline BMI: 27.24 ± 5.37).
Thirty metabolites were significantly associated with change in BMI per year in TwinsUK using Bonferroni correction. Four were independently associated with weight change in the multivariable LASSO model and replicated in KORA: namely, urate (meta-analysis β [95% CI] = 0.05 [0.040 to 0.063]; P = 1.37 × 10 ), gamma-glutamyl valine (β [95% CI] = 0.06 [0.046 to 0.070]; P = 1.23 × 10 ), butyrylcarnitine (β [95% CI] = 0.04 [0.028 to 0.051]; P = 6.72 × 10 ), and 3-phenylpropionate (β [95% CI] = -0.03 [-0.041 to -0.019]; P = 9.8 × 10 ), all involved in oxidative stress. Higher levels of urate at baseline were associated with weight gain in KORA and PLCO.
Metabolites linked to higher oxidative stress are associated with increased long-term weight gain.
研究长期体重变化与血液代谢物之间的关联。
对来自英国双胞胎队列的3176名女性(年龄范围:18.3 - 79.6岁,基线体重指数:25.11±4.35)在8.6±3.79年期间的体重指数变化进行评估,随访时检测了280种代谢物。将具有统计学显著性的代谢物(校正协变量后)纳入多变量最小绝对收缩和选择算子(LASSO)模型。研究结果在奥格斯堡地区合作健康研究(KORA)中得到重复验证(n = 1760;年龄范围:25 - 70岁,基线体重指数:27.72±4.53)。该研究还检验了所识别出的代谢物是否能前瞻性地预测KORA研究以及前列腺、肺、结肠和卵巢癌(PLCO)研究(n = 471;年龄范围:55 - 74岁,基线体重指数:27.24±5.37)中的体重变化。
在英国双胞胎队列中,经Bonferroni校正后,有30种代谢物与每年体重指数的变化显著相关。在多变量LASSO模型中有4种代谢物与体重变化独立相关,并在KORA研究中得到重复验证,分别是尿酸盐(荟萃分析β[95%置信区间]=0.05[0.040至0.063];P = 1.37×10)、γ-谷氨酰缬氨酸(β[95%置信区间]=0.06[0.046至0.070];P = 1.23×10)、丁酰肉碱(β[95%置信区间]=0.04[0.028至0.051];P = 6.72×10)和3-苯丙酸(β[95%置信区间]= -0.03[-0.041至-0.019];P = 9.8×10),所有这些都与氧化应激有关。基线尿酸盐水平较高与KORA研究和PLCO研究中的体重增加有关。
与较高氧化应激相关的代谢物与长期体重增加有关。