Paternity leave duration and maternal depressive symptoms: a longitudinal study of paternal involvement and coparenting in Japan.
Paternity leave may be associated with maternal mental health through fathers' involvement and coparenting, but short-term family-process associations during the early postpartum period remain unclear.
We examined modeled structural and indirect associations among paternity leave duration, mother-reported paternal involvement and coparenting, and maternal depressive symptoms from 3 to 6 months postpartum, separately by parity.
This two-wave online survey included 396 Japanese mothers (244 primiparas and 152 multiparas). At the 6-month follow-up, mothers reported the father's total paternity leave duration, categorized as no leave, short leave (1-31 days), or long leave (32 days or more). Paternal involvement at 3 and 6 months was modeled as a latent construct indicated by four domains of childcare and housework; coparenting and depressive symptoms were assessed using the CRS-J and CES-D, respectively. A multigroup two-wave structural equation model of structural and indirect associations was estimated. Indirect associations were evaluated using 5,000 bootstrap samples and bias-corrected 95% confidence intervals. All principal variables were mother-reported.
The freely estimated multigroup model showed good fit, χ2(194) = 309.282, CFI = 0.961, TLI = 0.946, and RMSEA = 0.039. Long leave, relative to no leave, was associated with greater paternal involvement at 3 months in both parity groups, whereas short leave was not clearly associated with paternal involvement. Paternal involvement and coparenting showed continuity from 3 to 6 months. The CRS-at-6-months to CES-D-at-6-months association and the total indirect association between long leave and CES-D were statistically supported within primiparas but not within multiparas. However, formal metric-invariant between-group comparisons were not significant for either the CRS-to-CES-D path, Δχ2(1) = 0.482, p = 0.488, or the total indirect association, ΔB = -0.225, 95% bias-corrected CI [-0.903, 0.441], p = 0.454.
Long paternity leave was associated with greater mother-reported paternal involvement and short-term coparenting continuity. The modeled indirect association with maternal depressive symptoms was very small, and formal comparisons did not establish parity differences. These findings describe short-term observational associations rather than causal effects.
We examined modeled structural and indirect associations among paternity leave duration, mother-reported paternal involvement and coparenting, and maternal depressive symptoms from 3 to 6 months postpartum, separately by parity.
This two-wave online survey included 396 Japanese mothers (244 primiparas and 152 multiparas). At the 6-month follow-up, mothers reported the father's total paternity leave duration, categorized as no leave, short leave (1-31 days), or long leave (32 days or more). Paternal involvement at 3 and 6 months was modeled as a latent construct indicated by four domains of childcare and housework; coparenting and depressive symptoms were assessed using the CRS-J and CES-D, respectively. A multigroup two-wave structural equation model of structural and indirect associations was estimated. Indirect associations were evaluated using 5,000 bootstrap samples and bias-corrected 95% confidence intervals. All principal variables were mother-reported.
The freely estimated multigroup model showed good fit, χ2(194) = 309.282, CFI = 0.961, TLI = 0.946, and RMSEA = 0.039. Long leave, relative to no leave, was associated with greater paternal involvement at 3 months in both parity groups, whereas short leave was not clearly associated with paternal involvement. Paternal involvement and coparenting showed continuity from 3 to 6 months. The CRS-at-6-months to CES-D-at-6-months association and the total indirect association between long leave and CES-D were statistically supported within primiparas but not within multiparas. However, formal metric-invariant between-group comparisons were not significant for either the CRS-to-CES-D path, Δχ2(1) = 0.482, p = 0.488, or the total indirect association, ΔB = -0.225, 95% bias-corrected CI [-0.903, 0.441], p = 0.454.
Long paternity leave was associated with greater mother-reported paternal involvement and short-term coparenting continuity. The modeled indirect association with maternal depressive symptoms was very small, and formal comparisons did not establish parity differences. These findings describe short-term observational associations rather than causal effects.