Objective To study the effect of family type on depression symptoms among high school students in Jiangsu Province and to provide evidence for mental health interventions among students.

Methods A self-administered anonymous electronic questionnaire survey was conducted from September to November 2020 among 22 076 high school students recruited from 13 prefectures in Jiangsu province using stratified multistage cluster random sampling. The Center for Epidemiologic Studies Depression Scale (CES-D) and the Internet Addiction Test (IAT) were used to assess students' mental health status and Internet addiction. The propensity score matching method was used to balance potential confounders of depression symptoms between two groups of students from normal (nuclear/extended) and abnormal (single-parent/reconstituted/grandparent-grandchild) families, and a total of 4 027 pairs (8 054) of students were matched with balanced confounders. A univariate logistic regression analysis model was used to examine the association between family type and depression symptoms among students.

Results Among all students, 26.1% were found to have symptoms of depression. The detection rate of depression symptoms among the 1 : 1 propensity score matched students was 27.3% for 4 027 individuals from normal families, with subgroup detection rates of 27.4%/27.1% for urban/suburban students and 29.7%/25.6% for boarding/day students, whereas the detection rate was 32.6% for 4 027 individuals from abnormal families, with subgroup rates of 32.9%/32.2% for urban/suburban students and 33.8%/31.9% for boarding/day students. The results of univariate logistic analysis of the data after propensity score matching indicated that students from abnormal families had an increased risk of depression symptoms compared with those from normal families (odds ratio OR = 1.29, 95% confidence interval CI: 1.17 – 1. 42); further subgroup analysis showed that the increased risk of depression symptoms was significant among students from abnormal families living in both urban (OR = 1.30, 95%CI: 1.15 – 1.48) and suburban (OR = 1.28, 95%CI: 1.10 – 1.48) areas, and among boarding students (OR = 1.20, 95%CI: 1.04 – 1.40) and day students (OR = 1.36, 95%CI: 1.20 – 1.54), respectively.

Conclusion Family type has a significant effect on the occurrence of depression symptoms among high school students in Jiangsu province. More attention should be paid to high school students living in abnormal family situations.