The performance of marginal structural models for estimating risk differences and relative risks using weighted univariate generalized linear models
Austin PC. Stat Methods Med Res. 2024; Apr 24 [Epub ahead of print].
We examined the risk of any perinatal mental illness associated with prepregnancy diabetes and identified how diabetes duration, complexity, and intensity of care affect this risk. We performed a population-based study of women aged 15–49 years with (n = 14,186) and without (n = 843,818) prepregnancy diabetes who had a singleton livebirth (Ontario, Canada, 2005–2015) and no recent mental illness. Modified Poisson regression estimated perinatal mental illness risk between conception and 1 year postpartum in women with versus without diabetes and in diabetes groups, defined by a latent class analysis of diabetes duration, complexity, and intensity-of-care variables, versus women without diabetes. Women with diabetes were more likely than those without to develop perinatal mental illness (18.1% vs. 16.0%; adjusted relative risk = 1.11, 95% confidence interval: 1.07, 1.15). Latent classes of women with diabetes were: uncomplicated and not receiving regular care (59.7%); complicated, with longstanding diabetes, and receiving regular care (16.4%); and recently diagnosed, with comorbidities, and receiving regular care (23.9%). Perinatal mental illness risk was elevated in all classes versus women without diabetes (adjusted relative risks: 1.09–1.12), but results for class 2 were nonsignificant after adjustment. Women with diabetes could benefit from preconception and perinatal strategies to reduce their mental illness risk.
Brown HK, Cairncross ZF, Lipscombe LL, Wilton AS, Dennis CL, Ray JG, Guttmann A, Vigod SN. Am J Epidemiol. 2020; 189(6):573-82. Epub 2019 Nov 12.
The ICES website uses cookies. If that’s okay with you, keep on browsing, or learn more about our Privacy Policy.