Premium
The Risk of Gestational Diabetes Mellitus Following Bereavement: A Cohort Study from D enmark and S weden
Author(s) -
László Krisztina D.,
Olsen Jørn,
Li Jiong,
Persson Martina,
Vestergaard Mogens,
Svensson Tobias,
Obel Carsten,
Cnattingius Sven
Publication year - 2015
Publication title -
paediatric and perinatal epidemiology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.667
H-Index - 88
eISSN - 1365-3016
pISSN - 0269-5022
DOI - 10.1111/ppe.12198
Subject(s) - medicine , gestational diabetes , odds ratio , pregnancy , relative risk , obstetrics , confounding , confidence interval , diabetes mellitus , sibling , population , cohort study , cohort , gestation , endocrinology , environmental health , developmental psychology , psychology , genetics , biology
Background We investigated whether bereavement of a close family member – a source of severe psychological stress exposure – the year before pregnancy is associated with gestational diabetes mellitus ( GDM ). Methods We studied pregnant women with livebirths in D enmark during 1994–2008 and with births in S weden during 1987–2006 ( n = 2 569 446). We obtained data on death of women's parents, siblings, and older children, and on demographic and health‐ and pregnancy‐related factors through linkage between nationwide registers. Results In multivariable models, death of a close relative the year before pregnancy was associated with a 14% increased odds of GDM [95% confidence intervals ( CI s) 1.03, 1.26]. The odds ratios corresponding to the loss of a child, parent, and sibling were 1.51 (95% CI : 1.17, 1.95), 1.12 (95% CI : 1.00, 1.25), and 0.68 (95% CI : 0.40, 1.25), respectively. Deaths due to cardiovascular diseases or diabetes were more closely related to the risk of GDM than other types of deaths. We found no association between unnatural deaths and the risk of GDM . Conclusions Death of a close relative the year before pregnancy was associated with a modestly increased GDM risk. Our findings according to the relative's cause of death suggest that differences in screening for GDM among exposure groups and residual confounding by familial factors related to metabolic and cardiovascular diseases may have contributed to this association. If there is a causal stress effect on GDM in this predominantly Nordic population, it is most likely small.