Physical growth and neurodevelopment during the first year of life: a cohort study of the Japan Environment and Children’s Study

The JECS is a nationwide, multicenter, prospective birth cohort study funded by the Ministry of Environment, Japan. The study design details have been described elsewhere [23, 24]. In brief, pregnant participants were registered between January 2011 and March 2014 in 15 regional centers covering a wide geographical area in Japan. During pregnancy, data were obtained during the first and second/third trimesters using self-administered questionnaires. Detailed information regarding the mother and child was obtained from medical records transcripts during the first trimester, at the time of delivery, and when the child was 1 month old. After delivery, data were collected at one and six months old and every six months until the child was six years old, then twice a year thereafter via self-reported questionnaires completed by the parents.

The JECS protocol was reviewed and approved by the Ministry of Environment’s Institutional Review Board for Epidemiological Studies and by the ethics committees of all participating institutions (#15000141). The ethical approval for this study was an extension of the ethical approval for the JECS protocol. Written informed consent was obtained from all parents.

In this study, we used the fixed dataset “jecs-an-20180131” released in March 2018. The dataset includes all available data of 104,065 fetuses, linked to their mothers’ data, collected until children were 12 months old. As subjects, we selected 89,953 children with live singleton term birth (≥ 37 and < 42 gestational weeks) who had parents of Japanese nationality and complete basic information on mother’s parity, child’s sex, length and weight at birth (Fig. 1). After excluding 3,148 children who had malformation/severe diseases, we eventually analyzed 86,805 children, including 44,264 boys and 42,541 girls.

The exposure variable was body growth from conception to 10 months old. Length and weight at birth, which can be viewed as growth from conception (length and weight = 0) to birth, were obtained from medical records. Subsequent anthropometric data were all derived from the personal health record of the child. Nearly all Japanese children receive a public health checkup at four months old. Their length and weight were thus directly measured by a public health nurse or a hospital nurse and officially documented in the record. They also receive similar checkups at 7 or 10 months old, occasionally at both 7 and 10 months old, depending on residency. Parents were asked to transcribe these anthropometric data at the 4-month checkup in the 6-month questionnaire and those at the 7- and 10-month checkups in the 12-month questionnaire. Because the data were not measured at exactly 4, 7 and 10 months, the values were extrapolated from the anthropometric data at birth and those at the 4-, 7- and 10-month checkups [25]. The extrapolated data were referenced to the Japanese growth standards separately by sex and then converted into z-scores (i.e. length-for-age, weight-for-age) [26, 27]. In the standards, the mother’s parity and child’s gestational age are additionally taken into consideration when calculating z-scores at birth.

These z-scored length and weight values at different ages were processed using conditional growth modeling, as in the studies conducted by Adair et al. and Horta et al. [19, 20] but additionally including birth length in our study. Conditional variables were obtained by regressing the current size (i.e. length or weight) against all previous sizes [21, 22]. At each time point, a conditional variable represents physical growth during a time interval, and a positive value represents a linear growth or relative weight gain faster than predicted in that period. For example, the conditional length at seven months old indicates the linear growth from four to seven months old. To estimate the conditional length, the current length was regressed against all previous lengths and weights. The conditional length at seven months old was derived by regressing the length at seven months old against the length and weight at four months old and birth. To estimate the conditional relative weight, the current weight was regressed against the current length and all previous lengths and weights. For instance, the conditional relative weight at seven months old was calculated by regressing the weight at seven months old against the length at seven months old and the length and weight at four months old and birth. Note that the length and weight at birth per se were treated as conditional variables at birth, as they had no previous anthropometric data to be regressed against. Consequently, our conditional variables included length and weight at birth and conditional length and relative weight at 4, 7 and/or 10 months old.

The outcome measure was neurodevelopmental delay at 6 and 12 months old using the Japanese-translated version of the Ages & Stages Questionnaires: third edition (ASQ-3). This version was prepared through a back-translation procedure and was approved by the publisher of the original English version [28]. The ASQ-3 can identify infants or young children who require further neurodevelopmental assessments to determine their eligibility for early intervention. The findings of the questionnaire basically agree with those of professionally administered developmental batteries [29, 30]. It has been used in clinical and research settings and translated into several languages [31‐33].

The ASQ-3 assesses five developmental domains: Communication, Gross Motor, Fine Motor, Problem Solving, and Personal-Social domains. For each domain, six skills are described to which parents answer “yes,” “sometimes,” or “not yet,” depending on whether or not their child has demonstrated the described skill. The responses are converted to points, with “yes” receiving 10 points; “sometimes”, 5 points; and “not yet”, 0 points. The child’s score for each developmental domain is the sum of all points received for the items under that domain and ranges from 0 to 60 points. The cut-off score for each domain is defined as two standard deviations below the mean score of large, standardized samples in the United States of America. A child is defined as having a neurodevelopmental delay if a score is at or below the cut-off level in any developmental domain. Although preliminary cut-off scores of the Japanese translation were recently proposed [34], these were not recommended to be used with confidence before 24 months old because of very limited sample sizes. Therefore, the cut-off scores were determined using the same methodologies as the original version, based on available data at 6 (n = 82,994) and 12 months old (n = 78,947) (Fig. 1), which represent the general Japanese population.

To examine which component and periods of physical growth are specifically associated with infant neurodevelopment, we conducted multivariable regression analyses using conditional variables at different ages. Because these variables are not correlated with each other, the variables and covariates were included together in the same model without concerns about collinearity [20]. The covariates were (i) mother’s age; (ii) maternal smoking during pregnancy, as recorded in the first trimester; (iii) maternal and (iv) paternal education level (junior high school, high school, and university or graduate school); (v) annual family income (< 4 000 000, 4 000 000–5 999 999, ≥ 6 000 000 JPY); (vi) home speech stimulation at 1 month old (whether a mother did or did not talk to her baby habitually). The ‘home speech stimulation’ covariate was used instead of the Home Observation for Measurement of the Environment scale [35], which is not employed in the JECS.

All analyses were performed using the R software program, version 4.0.3 (The R Foundation, Vienna, Austria). Quasi-Poisson regression models were used to estimate adjusted relative risks (aRRs) with 95 % confidence intervals (CIs).