Premium
Link between insomnia and perinatal depressive symptoms: A meta‐analysis
Author(s) -
Emamian Farnoosh,
Khazaie Habibolah,
Okun Michele L.,
Tahmasian Masoud,
Sepehry Amir A.
Publication year - 2019
Publication title -
journal of sleep research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.297
H-Index - 117
eISSN - 1365-2869
pISSN - 0962-1105
DOI - 10.1111/jsr.12858
Subject(s) - meta analysis , confidence interval , medicine , sample size determination , depression (economics) , categorical variable , breastfeeding , random effects model , demography , postpartum depression , statistics , pediatrics , pregnancy , mathematics , genetics , sociology , biology , economics , macroeconomics
Evidence shows the possible link between insomnia and perinatal depressive symptoms. In order to find a convergent quantitative answer, we collected data via the search of Medline, EMBASE and reference tracking, which included nine studies (a total sample of 1,922 women). An aggregate effect size estimate (correlation coefficient) was generated using the comprehensive meta‐analysis software. For the meta‐analytic procedure, a random effects model was set a priori . Moderating factors, including study design, method of assessment of depression, geographical origin of data, publication year, mean age, % married, breastfeeding rate, quality and type of data, % primiparous and history of depression, were examined via categorical or univariate mixed‐effects (method of moments) meta‐regression methods. Heterogeneity and publication bias were examined using standard meta‐analytic approaches. We found a significant, medium‐size relationship between insomnia and perinatal depressive symptoms (point estimate, 0.366; 95% confidence interval [CI], 0.205–0.508; p < 0.001; n = 9) and this was significantly heterogeneous ( Q , 118.77; df , 8; p < 0.001; I 2 , 93.26%). The effect size estimate was significant for studies reporting no history of depression (point estimate, 0.364; 95% CI, 0.035–0.622; p < 0.05; n = 5) and for study design. With meta‐regression, no moderating factor (age, marriage rate, breastfeeding rate, pregnancy history or publication year) significantly mediated the effect size estimate. The depression assessment scale used, but not other categorical variables, explained the magnitude of heterogeneity. We found that insomnia during the perinatal period is associated with depressive symptoms, which warrants screening pregnant mothers for insomnia and depression.