Xu Xiying, Qi Chao, Xu Shuang, Fu Xinhao, Li Zhiyuan, Ren Hong, Qian Qian, Guo Shanshan
Affiliated Hospital of Shandong University of Traditional Chinese Medicine, Jinan, China.
Affiliated Hospital of Shandong Academy of Traditional Chinese Medicine, Jinan, China.
J Affect Disord. 2024 May 1;352:259-266. doi: 10.1016/j.jad.2024.02.053. Epub 2024 Feb 15.
Recent years have seen increasing attention to improving depressive symptoms through dietary intakes, yet the association between thiamine intake and depression remains unclear. The present study aimed to explore this association using data from an American cross-sectional study.
We explored the association of covariates, exposure, and outcome with logistic regression equations. Multivariable regression models were performed to further exclude confounding factors. To investigate nonlinear relationships, we employed restricted cubic splines. Recursive algorithms were utilized to identify inflection points. Additionally, we conducted stratified analyses by age and sex to uncover differences among subgroups.
When all covariates were adjusted, the association between thiamine intake and depression was not statistically significant [0.93 (0.82, 1.07)]. In the linear trend test using Q1 as the reference, the ORs (95%CI) for Q2, Q3, and Q4 were 0.87 (0.73, 1.04), 0.83 (0.68, 1.00), and 0.92 (0.73, 1.16), which suggested that the association might be nonlinear. We then confirmed this nonlinear relationship with a restricted cubic spline, and the inflection point of 1.35 mg/day was calculated. Before the inflection point, the effect value of the relationship was 0.68 (0.53, 0.89). After the inflection point, no significant association was found [1.10 (0.92, 1.31)]. Stratified analyses revealed that this nonlinear relationship was consistent among women and individuals aged <60 years.
In this cross-sectional study among American general adults, we found a nonlinear association between thiamine intake and depression and further observed differences by age and sex.
近年来，通过饮食摄入改善抑郁症状受到越来越多的关注，但硫胺素摄入量与抑郁症之间的关联仍不明确。本研究旨在利用一项美国横断面研究的数据探讨这种关联。
我们使用逻辑回归方程探讨协变量、暴露因素和结局之间的关联。进行多变量回归模型以进一步排除混杂因素。为了研究非线性关系，我们采用了受限立方样条。利用递归算法识别拐点。此外，我们按年龄和性别进行分层分析，以发现亚组之间的差异。
当对所有协变量进行调整后，硫胺素摄入量与抑郁症之间的关联无统计学意义[0.93(0.82,1.07)]。在线性趋势检验中，以第一四分位数(Q1)为参照，第二四分位数(Q2)、第三四分位数(Q3)和第四四分位数(Q4)的比值比(ORs)(95%置信区间)分别为0.87(0.73,1.04)、0.83(0.68,1.00)和0.92(0.73,1.16)，这表明该关联可能是非线性的。然后我们用受限立方样条证实了这种非线性关系，并计算出拐点为1.35毫克/天。在拐点之前，该关系的效应值为0.68(0.53,0.89)。在拐点之后，未发现显著关联[1.10(0.92,1.31)]。分层分析显示，这种非线性关系在女性和60岁以下个体中是一致的。
在这项针对美国普通成年人的横断面研究中，我们发现硫胺素摄入量与抑郁症之间存在非线性关联，并进一步观察到了年龄和性别的差异。
