Koyama Y, Isumi A, Fujiwara T
Department of Public Health, Institute of Science Tokyo, Tokyo, Japan.
Department of Social and Behavioral Sciences, Harvard T.H. Chan School of Public Health, Boston, MA, USA.
Epidemiol Psychiatr Sci. 2025 Mar 26;34:e22. doi: 10.1017/S2045796025000162.
Mental health problems in adolescence are increasingly prevalent and have tremendous impacts on life-long health and mortality. Although household poverty is a known risk factor for adolescent mental health, evidence of the timing hypothesis is scarce. We aimed to examine the longitudinal associations of poverty across childhood with mental health in adolescence, focusing on the timing of exposure.
We used the data of 5,671 children from a Japanese population-based longitudinal cohort, which recruited the first graders (aged 6-7 years) and followed biannually until eighth grade (aged 13-14 years) in Adachi, Tokyo. Household poverty was defined as households having any of the following experiences: annual income less than Japanese yen 3 million, payment difficulties and material deprivations, measured in first, second, fourth, sixth and eighth grades. Adolescent mental health included parent-report internalizing and externalizing problems (the Strengths and Difficulties Questionnaire), self-report depression (the Patient Health Questionnaire-9) and self-esteem (the Japanese version Children's Perceived Competence Scale) in eighth grade. We applied g-estimation of structural nested mean modelling to account for time-varying confounders.
If adolescents were exposed to household poverty at any grade across childhood, on average, they would report more severe depressive symptoms (ψ = 0.32 [95% CI 0.13; 0.51]) and lower self-esteem (ψ = -0.41 [-0.62; -0.21]) in eighth grade. There were also average associations of household poverty at any grade with more internalizing (ψ = 0.19 [0.10; 0.29]) and externalizing problems (ψ = 0.10 [0.002; 0.19]). Although the associations between household poverty and mental health were stronger in younger ages (e.g., poverty in the second grade → depression: ψ = 0.54 [-0.12; 1.19] vs. poverty in the eighth grade → depression: ψ = -0.01 [-0.66; 0.64]), overlapping 95% CIs indicated no statistically significantly different associations by the timing of exposure.
We found the average effect of exposure to household poverty at any grade on mental health outcomes in eighth grade, failing to support the timing hypothesis. The findings indicate that the effects of household poverty accumulate over time in childhood and impact adolescent mental health (cumulative hypothesis) rather than the effects differ by the timing of exposure. While cumulative effects suggest a persistent intervention in poor households across childhood, we highlight intervention at any timing in childhood may be effective in alleviating adolescent mental health problems.
青少年心理健康问题日益普遍，对终身健康和死亡率有巨大影响。虽然家庭贫困是青少年心理健康的已知风险因素，但关于时机假说的证据很少。我们旨在研究童年时期贫困与青少年心理健康之间的纵向关联，重点关注暴露时机。
我们使用了来自日本一个基于人群的纵向队列的5671名儿童的数据，该队列招募了一年级学生（6 - 7岁），并在东京足立区每半年随访一次，直至八年级（13 - 14岁）。家庭贫困定义为家庭有以下任何一种经历：年收入低于300万日元、支付困难和物质匮乏，在一年级、二年级、四年级、六年级和八年级进行测量。青少年心理健康包括八年级时家长报告的内化和外化问题（长处和困难问卷）、自我报告的抑郁（患者健康问卷 - 9）和自尊（日本版儿童感知能力量表）。我们应用结构嵌套均值模型的g估计来考虑随时间变化的混杂因素。
如果青少年在童年的任何一个年级暴露于家庭贫困，平均而言，他们在八年级会报告更严重的抑郁症状（ψ = 0.32 [95% CI 0.13; 0.51]）和更低的自尊（ψ = -0.41 [-0.62; -0.21]）。家庭贫困在任何一个年级与更多的内化问题（ψ = 0.19 [0.10; 0.29]）和外化问题（ψ = 0.10 [0.002; 0.19]）也存在平均关联。虽然家庭贫困与心理健康之间的关联在较年轻的年龄段更强（例如，二年级贫困→抑郁：ψ = 0.54 [-0.12; 1.19] 与八年级贫困→抑郁：ψ = -0.01 [-0.66; 0.64]），但重叠的95%置信区间表明暴露时机的关联在统计学上没有显著差异。
我们发现童年任何一个年级暴露于家庭贫困对八年级心理健康结果的平均影响，但未支持时机假说。研究结果表明，家庭贫困的影响在童年时期随时间累积并影响青少年心理健康（累积假说），而不是影响因暴露时机而异。虽然累积效应表明需要对贫困家庭在整个童年时期进行持续干预，但我们强调在童年的任何时候进行干预都可能有效缓解青少年心理健康问题。
