The sociodemographic characteristics of youth and their families examined in ToyBox and HGS, respectively, are presented in Table 1. The ToyBox study examined a total sample of 1778 preschool children (4.9 ± 0.3 years old), while the HGS sample comprised 2294 preadolescent children (11.1 ± 0.7 years old). The samples were almost equally split into boys and girls. Almost two thirds of fathers (66.7%) and mothers (62.6%) of ToyBox participants were younger than 42 and 38 years of age, respectively. The opposite was observed for fathers and mothers of HGS participants, the majority of whom were older than 42 (62%) and 38 (60.8%) years of age, respectively. In both studies, more than 85% of children were Greek nationals, while parents of ToyBox participants had a higher educational level of more than 14 years (42.2% fathers and 49.5% mothers), compared to parents of HGS participants (34% fathers and 37.4% mothers). Furthermore, Table 1 also presents data on parental weight status, with 69.9% and 30.1% of fathers and mothers of preschool children in the ToyBox study being overweight or obese. Higher percentages were observed for parents in the HGS, where 74.6% and 39.9% of fathers and mothers of preadolescents were overweight or obese.

Table 2 presents the perinatal characteristics of participants. Pre-pregnancy weight status of mothers was similar in both studies, with 6.5% and 6.8% of mothers in ToyBox and HGS classified as underweight prior to their pregnancy, 15.2% and 14.3% of mothers categorised as overweight, and 4.2% and 4.1% classified as obese, respectively. Almost equal percentages were also observed for the weight gain of mothers during pregnancy according to the IOM (Institute of Medicine) recommendations. In this regard, 29% and 35.3% of mothers in ToyBox and HGS gained weight during their pregnancy that was below IOM’s recommendations, while for 36.2% and 32.3% of mothers in ToyBox and HGS, gestational weight gain was above IOM’s recommendations. Furthermore, more than 80% of children in both studies were born full term (i.e., ≥ 37 weeks of gestation). Regarding size at birth, 7.6% and 12.1% of preschool and preadolescent children in ToyBox and HGS were born SGA, while 7.4% of preadolescents in HGS were born LGA, compared to only 0.9% of preschool children born LGA in ToyBox. In terms of children’s weight gain and breastfeeding practices during the first six months of life, the relevant percentages were almost equivalent in both studies. In this regard, 31.2% of preschool children in ToyBox and 33% of preadolescents in HGS had rapid weight gain from birth until their sixth month of life, while 10.4% and 10.6% of preschool children and preadolescents, respectively, had poor weight gain. In terms of breastfeeding, only 10.1% of preschool children in ToyBox and 8.2% of preadolescents in HGS were exclusively breastfed as infants, until their sixth month of age. Finally, the vast majority of children in both studies (i.e., 72.3% in ToyBox and 66.8% in HGS) had solid food introduced into their diet at 5 to 6 months of age.

The percentages of children in the different weight status categories, both as a total and by sex, are presented in Fig. 1a for ToyBox and Fig. 1b for HGS. The ToyBox study sample had the higher proportion of children who were of normal weight status, i.e., 71.5%, in comparison to 55% in the HGS. On the contrary, the prevalence rates of overweight and obesity were much higher among preadolescents in the HGS (i.e., 30.5% and 11.6% respectively) compared to preschool children in the ToyBox study (i.e., 15% and 5.5% respectively). When stratifying by sex, the prevalence of overweight and obesity was found to be higher in preschool girls compared to preschool boys in ToyBox (17.4% and 7% for girls vs. 12.7% and 4.1% for boys; P < 0.05). The opposite was observed in the HGS, where the prevalence of overweight and obesity was found to be higher in preadolescent boys compared to preadolescent girls (31.3% and 13.7% for boys vs. 29.7% and 9.5% for girls; P < 0.05).

a ToyBox study, b HGS. *, †Percentages sharing the same superscript symbol are statistically significantly different between boys and girls (P < 0.05).

The associations between growth rate during the first six months of life and development of overweight or obesity by preschool and preadolescent years are presented as odds ratios (OR) (95% CI) in Table 3. The observed associations were found to be statistically significant only for preadolescents participating in the HGS, showing a higher likelihood of overweight/obesity for infants who had rapid weight gain in their first six months of life (OR: 1.36, 95% CI: 1.14–1.64 for the total population of preadolescents and OR: 1.76, 95% CI: 1.35–2.29 for the subsample of boys in the HGS). A statistically significant association was also observed between poor growth rate and development of overweight/ obesity for preadolescent boys in the HGS (OR: 1.51, 95% CI:1.03–2.20). No statistically significant associations were observed for preschool-aged children in ToyBox.

The results from the ROC analyses conducted for the identification of the optimal cut-off point values both for the changes of the examined growth indices from birth to six months of age, as well as for the values of these growth indices at six months of age that were associated with the lower likelihood of overweight/obesity with the highest possible sensitivity (Se) and specificity (Sp) in preschool and preadolescent years are presented in Tables 4, 5, respectively. Based on the data presented in Table 4, changes in WAZ, WLZ, and BAZ by 0.54 (Se: 53.8%; Sp: 54.4%), 0.84 (Se: 62.4%; Sp: 51.3%) and 0.82 (Se: 49%; Sp: 59.8%), respectively, were significantly associated with a lower likelihood for the occurrence of overweight or obesity in preschool years. Similar findings were also observed for preadolescents, since changes in WAZ, WLZ, and BAZ by −0.91 (Se: 48%; Sp: 57.8%), 0.89 (Se: 52.9%; Sp: 57.1%) and 0.77 (Se: 53.1%; Sp: 56.6%), respectively, were significantly associated with a lower likelihood for the occurrence of overweight or obesity in preadolescence. No statistically significant findings were observed for the change in HAZ in either study.

According to the results presented in Table 5, values of WAZ, WLZ, HAZ and BAZ at 6 months of age up to 0.58 (Se: 55.1%; Sp: 71.2%), 0.10 (Se: 54.2%; Sp: 68.4%), 0.93 (Se: 57.3%; Sp: 52.3%) and −0.005 (Se: 53.6%; Sp: 69%), respectively, were significantly associated with a lower likelihood for the occurrence of overweight or obesity in preschool years. Furthermore, values of WAZ, WLZ and BAZ at 6 months of age up to −0.95 (Se: 46.2%; Sp: 57.7%), 0.09 (Se: 50.7%; Sp: 61.1%) and 0.23 (Se: 52.3%; Sp: 59.5%), respectively, were significantly associated with a lower likelihood for the occurrence of overweight or obesity in preadolescence.