Association of Maternal Preeclampsia With Offspring Risks of Ischemic Heart Disease and Stroke in Nordic Countries

This cohort study examines the risk of cardiovascular diseases among offspring who had prenatal exposure to severe or mild forms of preeclampsia.


eAppendix 1. Description of the source, the measurement, and the categorization of covariates
Information on calendar year of birth (1973-1978, 1979-1984, 1985-1990, 1991-1996, 1997-2002, 2003-2008, 2009-2016), sex (boy or girl), gestational age, birth weight, and multiple/singleton birth was obtained from the Medical Birth Registers (MBR) in Denmark, Finland, and Sweden. Gestational age was based on routine ultrasound examinations performed in early second trimester or date of last menstrual period (LMP). Ultrasound examinations was not uniformly introduced until the mid-1990s in Denmark, Sweden and Finland. In the rare event of missing information on ultrasound examinations, we used LMP (if it was available) to estimate gestational age. We excluded offspring with abnormal gestational age (<154 or >315 days) or with abnormal birth weight (more than five standard deviations below or above the mean).
Diagnoses of congenital anomalies in study participants were extracted from the Danish National Patient Register (DNPR), the Swedish Patient Register (SPR) and the Finnish Hospital Discharge Register (FHDR) using the International Statistical Classification of Diseases and Related Health Problems (ICD) codes: ICD-8/9: 740-759 and ICD-10: Q00-Q99.
Maternal country of origin (same as the study country versus not) and marital status before delivery (married/registered partnership versus not) was identified from the Danish Civil Registration System, the Finnish Central Population Register, and the Swedish Total Population Register. Data on the mother's highest education before delivery was obtained from the Danish Integrated Database for Labour Market Research, the Education Register at Statistics Finland, and the Swedish Register of Education. From the MBR we also obtained data on maternal age at delivery (≤19, 20-24, 25-29, 30-34, and ≥35 years), parity (1,2, ≥3), body mass index (BMI, not available in our Finnish data) and smoking (yes or no) in early pregnancy. Maternal BMI was categorized as underweight (BMI <18.5 kg/m 2 ), normal weight (18.5-24.9 kg/m 2 ), overweight (25.0-29.9 kg/m 2 ), and obesity (≥30.0 kg/m 2 ).
Our criteria for selecting confounders were (1) a known or an a-priori hypothesized association with both exposure and outcome and (2) not being on the causal pathway between exposure and outcome.
The sibling analysis included all sibling pairs from our whole population. Firstly, we identified the mothers who had two or more biological children, and created a sub-cohort of these sibling pairs. Secondly, we performed analyses in this sibling sub-cohort using a stratified Cox regression with a separate stratum for each family identified by the mother's identification number. We adjusted for the same covariates as in the analyses involving the whole population. Only sibling pairs discordant for exposure and outcome contributed to the estimates in the analyses.
The sibling analysis included 7 031 134 sibling pairs with 2 808 293 mothers, among them 15 851 mothers had recurrent preeclampsia. The number of sibling pairs discordant for exposure to maternal preeclampsia and the diagnosis of ischemic heart disease or stroke in their lives was 290 027, 14 093, or 21 998, respectively.