Association between rapid weight gain in early years and subsequent adiposity indices in Portuguese children aged 3 to 5 years

The present study aimed to evaluate the association of rapid weight gain (RWG) in early years with subsequent measures of adiposity in Portuguese children. We used data from a cross-sectional study of a representative sample of 5359 Portuguese children aged 3 to 5 years in 2009-10. Measures of adiposity


Introduction
The levels of body mass index (BMI) and overweight/obesity have increased rapidly in recent decades across age groups and in most populations (NCD-RisC, 2017).The worldwide prevalence of overweight and obesity in children aged 0-5 years increased from 30.3 mil-lion to 38.9 million between 2000 and 2020 (FAO, 2022).In Portugal, in 2018/19, the prevalence of childhood overweight and obesity remained high at 29.7% and 11.9%, respectively (PORTUGAL, 2021).Childhood obesity tends to persist into adulthood (Reilly and Kelly, 2011) and is associated with a range of adverse health outcomes (NCD-RisC, 2017), in-indices included age-specific body mass index z-score (BAZ) and skinfold (suprailiac, subscapular and triceps) thickness.RWG was defined as the difference between z-scores of weightfor-age and birthweight > 0.67 Standard Deviation.Quantile regression models were used to estimate the difference in the 50 th (median) and 90 th (upper end) percentiles of each measure by RWG.Rapid weight gain in early years (62.5% of children) was associated with greater levels of BAZ and skinfold thickness measures, both for the median and 90 th percentile.For the median, the difference was 1.08 (95% CI: 1.01-1.16)for BAZ, 1.11 mm (95% CI: 0.98-1.25)for suprailiac, 0.99 mm (95% CI: 0.87-1.11)for subscapular and 1.34 mm (95% CI: 1.11-1.57)for triceps skinfolds thickness after the adjustments.The increases in the 90th percentiles of adiposity measures associated with RWG were greater than those in median levels.RWG in childhood was associated with increased levels of adiposity measures, independent of prenatal and childhood factors.
Early years are a key risk period for the development of obesity (Gillman, 2010), and there are growing studies confirming an association between early rapid postnatal growth and subsequent higher adiposity measures (Ong and Loos, 2006;Zheng et al., 2018).Recent studies have shown that rapid weight gain (RWG) in early years, defined as a change in weight-for-age > 0.67 standard deviation (Ong et al., 2000), is an important early life risk factor for the development of obesity in the short, medium and long term (Woo Baidal et al., 2016).A meta-analysis showed that children with RWG by two years of age were 3.7 times more likely to become obese in adulthood (Zheng et al., 2018).Furthermore, systematic reviews also found that RWG was related to increased body fat percentage, waist circumference, and waist-to-height ratio (Lucas, 2010).However, most of these studies tend to focus primarily on a single measure of adiposity such as BMI.
The first thousand days of the child's life are a crucial window for prevention and intervention measures for reducing the risk of obesity (Mameli et al., 2016).Therefore, this study aimed to examine the association of RWG in early years with BMI and three skinfold thickness measures in a nationally representative sample of Portuguese children aged three to five years.

Methods
We used the data from a cross-sectional survey, the "Portuguese Prevalence Study of Obesity in Childhood, changes from 2002 to 2007.The Effects of Lifestyle and Environmental Factors".The survey was conducted between March 2009 and January 2010 in public and private kindergartens and primary schools in all districts of Portugal, except the Portuguese archipelagos (Madeira and Azores).The study sample was selected using proportional stratified random sampling according to the number of children in each district.This is a nationally representative survey totaling 17.509 children aged 3 to 10 years.The participation rate was 57.4% (Jago et al., 2012).The study was approved by the Directorate-General for Innovation and Curriculum Development and all participants received information about the research objectives and signed the Informed Consent Form.

Study sample
In this cross-sectional study, children between the ages of 3 and 5 years were included.Children without birthweight, height or weight measurements, or with implausible values were excluded, i.e. standardized weight-for-age index (WAZ) scoring < -6 or > 5, and BMI for age (BAZ) scoring < -5 or > 5 (WHO, 2011).The final eligible sample included 5359 children (Figure 1).

Measures
Trained researchers using standardized procedures performed anthropometric measurements.Height and weight of children were measured to the nearest 0.1 cm and 100 g, with light clothing and without shoes, using a portable stadiometer and a portable electronic scale, respectively.The thickness of the suprailiac skinfold (SSF), subscapular skinfold (SSSF) and triceps skinfold (TSF) (mm) were measured twice with a skinfold caliper and the average of the two readings was used for each skinfold measure (Li et al., 2016).BMI (kg/m 2 ) was calculated for each child and converted to age-and genderspecific BMI z-score using the WHO Anthro Program (WHO, 2011).Birthweight was reported by parents based on data from the individual card that each child got from the maternity where they were born.In this card, they have all the information concerning the birth.

Exposure and outcome
Rapid weight gain (RWG) as an exposure with the definition of a change in weight-for-age z score > +0.67 (from birth to collection date) is the most frequent and widely accepted definition (Ong et al., 2000).The outcome measures were BMI z-score (BAZ) and three (suprailiac, subscapular and triceps) skinfolds thickness measures.

Confounders
Covariates included the mother's height and weight (for calculating maternal BMI), maternal age, maternal education (low -primary/sub-secondary ≤9 years, middle -secondary 10-12 years, and high -higher education based on the Portuguese Educational System), and duration of breastfeeding (months).

Statistical analysis
They were summarized using mean values (standard deviation) for continuous variables and frequencies for categorical variables.Their differences by RWG group were tested using t-student and chi-square tests, respectively.As the distributions of adiposity measures were skewed (Yu et al., 2003), we applied quantile regression models to examine whether RWG was associated with the median (50 th percentile) and upper end (90 th percentile) of the adiposity measures (Koenker, 2005).The differences (95% Confidence Interval) in median and 90 th percentiles for each adiposity measure between children with and without RWG were estimated.We first fitted an unadjusted model (Model 1), then ad-justed for maternal BMI, maternal age, and birthweight (Model 2), and further for maternal education and duration of breastfeeding (Model 3).Data were analyzed using the software STATA (v14).

Results
This study includes 5359 children from 3 to 5 years, being 50.46% (n = 2704) boys and 49.54% (n = 2655) girls.The children had a mean birth weight of 3207.87g.The children' s mothers had a mean age of 34 years.
There was a high proportion of children who experienced RWG (62.5%).They had a lower mean birthweight than those who did not.There was no difference in mean maternal BMI, maternal age, maternal education, and duration of  3).

Discussion
The present study showed that children with RWG had greater levels of adiposity of several adiposity indices (e.g., BAZ, SSF, SSSF, TSF), both in median and upper end of distributions after adjusting for prenatal and childhood factors.These differences were consistently greater in the 90 th than the 50 th percentile, indicat-breastfeeding between those with and without RWG (p>0.05)(Table 1).
RWG was associated with increased levels of all adiposity measures (BAZ, SSF, SSSF, TSF) at both the 50 th and 90 th percentiles.The differences in mean adiposity measure by the RWG group increased after adjusting for maternal BMI, maternal age, and birthweight, and maintained when further adjusting for maternal education and breastfeeding duration (Figures 2 and 3).The adjusted difference (model 3) in median BAZ was 1.08 z-score (95% CI: 1.01-1.16).The respective difference in skinfold thickness was 1.11 mm (0.98-1.25) for SSF, 0.99 mm (0.87-1.11) for SSSF, and 1.34 mm (1.11-1.57)for TSF (Figure 2).
RWG was also associated with the upper end of the distribution (90 th per- ing that RWG not only increased the median, but increased more for the extreme high levels of adiposity measures. A recent systematic review and meta-analysis of populations from countries on five continents showed that RWG by the age of 2 years was associated with increased risk of overweight in childhood and adulthood (Zheng et al., 2018).Our findings are consistent with previous studies (Nanri et al., 2017;Zheng et al., 2018).It has been reported that changes in velocity of weight growth in the early years have long-lasting effects for adipose tissue accumulation (Hooper et al., 2001;Godfrey et al., 2007) and other health outcomes (Godfrey et al., 2010;Hanley et al., 2010;Lillycrop, 2011;van der Haak et al., 2019).Using the same definition of 0.67 standard deviation of the WAZ, Brazilian children showed a 69% and 64% increase in the overweight and obesity risks, respectively (Monteiro et al., 2003).
In general, large infants tend to be at greater risk of overweight and obesity during adolescence (Lyons-Reid et al., 2021).However, accelerated growth in premature, small for gestational age, or low birthweight infants is associated with a range of long-term deleterious effects (Victora et al., 2001;Martin et al., 2017), including obesity and hypertension (Parsons et al., 1999).Infants born with low birthweight have a higher risk of having increased growth velocity (Gillman et al., 2010).Given that the relationships between extremes of weight (high or low) outcomes associated with adiposity in later life are not consistent, rapid weight gain in children may be an alternative ex- planation to this controversy.
Some studies have shown that exclusive breastfeeding longer than 6 months and delayed exposure to food formula may protect from increased adiposity throughout adolescence (Cordero et al., 2014).A cohort of Australian children showed that formula-fed children had more accelerated BMI growth trajectories compared to those who were breastfed (Oddy et al., 2014).In the present study, breastfeeding duration was not associated with BAZ or with skinfold thickness values, and the adjustment for breastfeeding did not alter the findings.Lack of information on food consumption patterns before or during breastfeeding (e.g., introduction of formula milk or other weaning foods) may explain the inconsistent findings from previous studies.
Health care in the first years of life provides a foundation for good health in later life, especially the period of the first 1000 days from conception to two years of age.In the context of an ecological model, maternal conditions before and during pregnancy and nutrition in early childhood are crucial to mitigate the burden of diseases in adulthood (Darnton-Hill et al., 2004).It is unclear which period of childhood is most critical for the development of obesity in children with RWG (Ong and Loos, 2006).Although evidence has suggested that the first two years are the period of greatest risk for the development of overweight (Zheng et al., 2018;Li et al., 2020), the age between 2 and 5 years was also indicated (Matos et al., 2011).Thus, this set of studies reinforces the need for careful monitoring of postnatal weight gain in these groups and referral to health professionals responsible for specific assessment and intervention.A comprehensive understanding of the importance of predictors of obesity in children is essential for the development and refinement of strategies aimed at improving maternal and child health care in childhood and other life cycles.
It is noteworthy that our findings are from a large, nationally representative sample of Portuguese children and include multiple measures objectively collected of subcutaneous adipose tissue.Although these measures are correlated, they indicate different aspects of adiposity (e.g., BAZ representing overall adiposity and skinfold thickness, which is a proxy for fat mass).Limitations include the cross-sectional design that does not allow us to assess temporality nor the possibility of exploring trajectories of the anthropometric measures measured.Thus, future investigations using longitudinal data are necessary.

Conclusion
RWG in the early years of life was associated with increased levels of adiposity in children aged 3 to 5 years, especially at the upper end of distributions, independent of prenatal and childhood factors.These results highlight the importance of monitoring the child's weight growth from early years as a potential preventive action for childhood obesity in primary care, as well as, to promote healthy child development and growth, and adulthood.
* p < 0.05 for difference between RWG and non-RWG groups (t-test for continuous variables and chisquare test for categorical variables).†Children (3 -5 years) with adiposity measures.