Population and study design
The BISCS was a representative, prospective cohort study conducted by China Medical University between 2017 and 2020. The design of the research has been previously published [17]. The BISCS was designed to investigate women and children’s health, particularly in development with psychiatric and neurological. In-person visits were conducted with mothers between 13-27 weeks and 28-36 weeks of gestation to record maternal demographic information and collect biological samples. The visits were conducted again at one month, three months, six months and twelve months postpartum to record maternal postnatal data and infant information, and use the Gesell Developmental Scale to assess the infant cognitive and neurodevelopmental levels six months and twelve months old. Children born to mothers before 36 weeks gestation, with severe complications during pregnancy, diseases in the nervous system around birth ,as part of a multiple pregnancies (e.g., twins, triplets), or without complete baseline and follow-up data were excluded. As Figure 1, a final cohort of 688 women was completed at delivery, six months postpartum (n = 459) , and twelve months postpartum (n = 449). There was no difference found in maternal and infant demographic information between attrite participants and those with complete information (see Table. S1).
Measures
Anthropometric outcomes
The weight and length at birth of the infants were recorded from the medical records, then at six months old (mean: 6.87 ± 1.27) and twelve months postpartum (mean: 11.67 ± 1.41) by BISCS staff, according to a standard protocol. The weight and length were measured with a digital scale and a stadiometer while children were wearing no shoes and light clothes (Seca 416 and 376 +; Seca corporation, Hamburg, Germany). Birth Z -score and Z-scored BMI for age and sex-specific weight for age (WFA) were calculated based on the World Health Organization (WHO) child growth reference [18]. Weight gain velocity was calculated as the change in WFA Z -score from birth to six months and six months to twelve months [19].
Infants’ development in neurological and psychiatric
Participants in this birth cohort were administered the Gesell Development Scale (GDS), which has been used extensively after being translated and standardized (Beijing Children's Health Care Institute, 1985) [20]. The Gesell Development Scale is used to evaluate the function of the central nervous system and identify defects in the neuromuscular or sensory system, which consists of five domains, including the adaptive domain (coordination, imitation, object recovery, discrimination and perception), gross motor domain (changes in posture, head balance and behavioral observations of standing, sitting and walking) and fine motor domain (ability to hold objects with fingers), language domain (vocabulary comprehension and dialogue skills), and social domain (social habits, reactions to persons, autonomy and independence). Developmental quotient (DQ) is defined as the quotient between developmental age and actual age, and the DQ of each child was calculated for each specific domain. Higher developmental quotient (DQ) scores mean higher cognitive levels [21]. Two well-trained pediatricians from the Shengjing Hospital of China Medical University assessed all participants to maximize reliability. They assessed all children in their assigned domain in order to avoid inter-examiner variability.
Covariates
Based on previous research [22-25], other variables were used to describe maternal and infants’ demographic characteristics and control for confounding variables. These variables were collected from hospital records (maternal age, maternal pre-pregnancy BMI [26], gestational weight gain (GWG) , delivery mode, parity, child sex, gestational weeks, and birth weight) and questionnaires (maternal education, annual family income, the primary caregiver, mode of infant feeding, the introduction of solid foods at six months and micronutrients supplementation (Iron, Vitamin B12 and Zinc) at six months). Only 0.85% of women smoked and 1.15% of women were drinking during pregnancy, so this variable was not included. We calculated the GWG rate (kg/week) as total GWG divided by the number of gestational weeks at delivery. Based on the IOM, above optimal weight gains are above 16 kg for mothers with normal weight and more than 11.5 kg for overweight mothers, respectively [27,28]. We categorized the mode of infant feeding as: formula feeding only, mixed breast and formula, and breast milk feeding only. Micronutrient supplementation (Iron, Vitamin B12 and Zinc) was treated as a dichotomized variable (yes/no). To ensure data integrity, wholly conditional specification multiple imputation (m=20) was used to impute missing data.
Statistical analyses
Longitudinal analyses
Multiple linear regression analyses were fitted to assess the association of infants’ BMI Z -score and weight gain velocity with the development in neurological and psychiatric status at the age of six and twelve months. The analyses used the repeated measured weight status and development in neurological and psychiatric variables at two-time points. The analyses were adjusted for all the above confounding factors. Furthermore, the associations between development in neurological and psychiatric status at six months and weight status at twelve months were additionally adjusted for weight status at baseline (six months) to study whether development in neurological and psychiatric status predicted the change in infants’ Z -scored BMI or weight gain velocity, and vice versa. Linear regression analyses were run using SPSS version 20.0 (SPSS 20.0, [2011], IBM).
Cross-lagged analyses
This cross-lagged analysis included confounders, stability effects, cross-sectional associations, longitudinal associations, and cross-lagged associations. Firstly, we examined the stability model, which only included the cross-sectional analysis, with confounding factors regressed at two baselines. Then, based on the first step, we entered the lagged association between infant BMI Z -score and weight gain velocity at six months with development in neurological and psychiatric status at twelve months, and the lagged association between development in neurological and psychiatric status at six months with infant BMI Z -score and weight gain velocity at twelve months. Finally, the above two lag associations entered the whole model at the same time. The above relationships were adjusted for all the above confounding factors. Besides, the cross-lagged associations were adjusted for weight status at baseline (six months) or development in neurological and psychiatric status at baseline (six months). The cross-lagged analyses were conducted with Mplus, version 7.11.