Parental Separation and Cardiometabolic Risk Factors in Late Adolescence: A Cross-Cohort Comparison

Abstract The aim of this study was to explore the association between parental separation during childhood (up to 18 years of age) and cardiometabolic risk factors (body mass index, fat mass index, blood pressure, physical activity, smoking, and alcohol consumption) in late adolescence using a cross-cohort comparison and to explore whether associations differ according to the age at which the parental separation occurred and the presence or absence of parental conflict prior to separation. Data from the Avon Longitudinal Study of Parents and Children (ALSPAC, United Kingdom) (1991–2011) and the 1993 Pelotas Birth Cohort (Brazil) (1993–2011) were used. The associations of parental separation with children's cardiometabolic risk factors were largely null. Higher odds of daily smoking were observed in both cohorts for those adolescents whose parents separated (for ALSPAC, odds ratio = 1.46; for Pelotas Birth Cohort, odds ratio = 1.98). Some additional associations were observed in the Pelotas Birth Cohort but were generally in the opposite direction to our a priori hypothesis: Parental separation was associated with lower blood pressure and fat mass index, and with more physical activity. No consistent differences were observed when analyses were stratified by child's age at parental separation or parental conflict.

with the father of the child) versus all other relationship statuses. Any divorce or separation reported after by either the mother or the child in multiple questionnaires was considered to indicate parental separation.
In the 1993 Pelotas Cohort, mothers were asked in the perinatal questionnaire if they lived with a partner or husband; as in ALSPAC a binary variable for cohabitation versus any other relationship status was used. Adolescents reported at age 18 whether their parents had separated, and this was used as the measure of parental separation in this cohort.

Details of assessment of blood pressure and physical activity
In ALSPAC, for both SBP and DBP the mean of two right arm measures was used.
PA was objectively measured using an Actigraph AM 7164 2.2 accelerometer, which the participants were requested to wear for 7 consecutive days during waking hours. Mean daily minutes of moderate-to-vigorous physical activity (MVPA) based on valid days was used 29 .
In Pelotas, for both SBP and DBP the mean of two left arm measures was used. PA was objectively measured using a GENEActiv accelerometer, which the participants were requested to wear 24 hours a day for 5 to 8 days. The mean MVPA over valid assessment periods of at least 10 minutes was used 30 .

Details of assessment of covariates included in the analyses
In ALSPAC, maternal age (years) was collected at 8 weeks gestation. Maternal education in 5 categories (CSE -Certificate of secondary education; Vocational degree; O levelordinary level; A leveladvanced level; University degree) was assessed at 32 weeks of pregnancy. Family income was measured at 33 and 47 months after delivery, and a mean family income was calculated; within-cohort quintiles of family income were generated from the full available sample. Information on ethnicity was reported at 12 years (white, black/mixed/other). Parental smoking was based on maternal and paternal self-report during pregnancy (8, 18 and 32 weeks) and 8 months after birth and those mothers and fathers who smoked at least one cigarette per day in any trimester of pregnancy or at 8 months after birth were classified as smokers. Relationship conflict was assessed based on four questions: (1) "How many arguments or disagreements have you and your partner had in the past three months? (none/ 1-3/ 4-7/ 8-13/ 14 or more)", "In the past 3 months, have any of these happened:" (2) "one of you walking out of the house", (3) "hitting or slapping partner", (4) "throwing or breaking things" (yes, I did this/ yes, he did this/ yes, we both did this/ no, not at all). A binary variable indicating relationship conflict was created, based on whether the couple had 8 or more disagreements or answered yes to any of the other questions.
Relationship conflict was assessed at three time points: infancy (assessed at 22 months), early childhood (assessed at 33 months), and mid childhood (assessed at 9 years). Associations for separation <5 years were adjusted for the infancy measure, separation 5-9 years were adjusted for the early childhood measure, and 10-18 years were adjusted for the mid childhood measure.
In Pelotas, maternal age (years), maternal schooling (0-4 years; 5-8 years; 9-11 years; 12+ years), family income of the month prior to the delivery (quintiles), and parental smoking status were assessed at the perinatal visit. Quintiles of family income were generated based on the full original sample. Parental smoking status was based on maternal self-report, and mothers and fathers who smoked at least one cigarette per day during any time of pregnancy were considered smokers. At 11 years old, skin color (white, non-white) and childreported parental conflict were assessed. Parental relationship conflict was assessed by the statement "You think the relationship between your father and your mother is:", having as response options: excellent, very good, good, average or bad. A binary variable for relationship conflict was created, with conflict assumed to be present when the adolescent classified the parents' relationship as average or bad (the two lowest categories). Analyses were repeated with only bad parental relationship (6.6%) included as a measure of parental conflict and stronger associations between parental divorce and cardiometabolic outcomes were found, but overall conclusions were unaltered.  (1991-2011) and 1993Pelotas Cohort (1993-2011.

Pelotas Cohort
Continuous outcomes: Mean differences compared to children whose parents did not separate ( (1991-2011) and 1993Pelotas Cohort (1993-2011, According to Parental Relationship Conflict.
Conflict before 5 y Conflict before 5 -9 y Conflict before 10 -18 y  DBP: diastolic blood pressure; PA: physical activity; MVPA: moderate-to-vigorous physical activity a BMI and FMI given as kg/m 2