Table 1.
Summary of background characteristics. (a) all infants, (b) infants with glabella-occipital protuberance (G-O), left tragion-right tragion (LT-RT), and head circumference (HC) measurements, excluding age groups other than 2, 3, 6, and 12 months, (c) the number of infants with G-O, LT-RT, and HC data in each age group and (d) the number of 3-month-old infants with G-O, LT-RT, HC data in each birth year group.
Fig 1.
Cephalic Index (CI), Globularity Index (GI) and head volume estimation.
Fig 2.
A pictorial explanation of the Globularity Index (GI).
Fig 3.
Trends observed in the direct measurements (left tragion-right tragion (LT-RT), glabella-occipital protuberance (G-O), and head circumference (HC)) over the first 12 months after birth. (a) LT-RT, (b) G-O, and (c) HC. Statistical differences across age groups (n = 207 for 2 months, n = 573 for 3 months, n = 100 for 6 months, and n = 29 for 12 months) were confirmed using factorial ANOVA with age group, sex and birth year as main effects, with the effect of age group on all direct measurements giving p < 0.001. The significance of pairwise differences (with lines connecting the pairs) between estimated marginal mean based on t-test and FDR correction are indicated as follows with asterisks: ***(p < 0.001). For more details on post hoc analysis, see S2-S4 Table in S1 File.
Fig 4.
Trends observed in the estimated measurements (half-width (a1), half-length (a2), height (h), CI, height/half-length ratio (h/a2), head volume, and GI) over the first 12 months after birth.
(a) a1, (b) a2, (c) h, (d) CI, (e) h/a2, (f) head volume, and (g) GI. Differences among the monthly segments (2 months: n = 207,3 months: n = 573, 6 months: n = 100, 12 months: n = 29) were statistically significant, as determined using a factorial ANOVA with age in months, sex and birth year as main effects; the effect of age on all estimated measurements adjusting for sex and birth year was significant, with p < 0.001 for all measurements except GI (p = 0.004). The significance of pairwise differences between estimated marginal mean (with lines connecting the pairs) based on t-test and FDR correction are indicated with asterisks as follows: *(p < 0.05), ***(p < 0.001) and none (not significant). See S6-S12 Tables in S1 File for more details on post hoc analysis.
Fig 5.
Scatter plot of CI versus birth year.
Fig 6.
Trends observed in the estimated measurements (half-width (a1), half-length (a2) and height (h)) at 3 months over the period from 2010 to 2019.
(a) a1, (b) a2, and (c) h. A letter was allocated to each birth year from left to right as indicated inside the box. The letter assigned to the group that came first in the sequence was added above the violin and box plots of the group that appear later in the sequence if the difference was significant at either p < 0.05 (in black), p < 0.01 (in blue), or p < 0.001 (in red) level based on the pairwise t-test with FDR correction conducted to contrast estimated marginal means. The number of infants was n = 52 for 2010, n = 100 for 2011, n = 148 for 2012, n = 25 for 2013, n = 52 for 2014, n = 15 for 2015, n = 19 for 2016, n = 51 for 2017, n = 63 for 2018, and n = 48 for 2019. See S24-S32 Tables in S1 File for details of post hoc analyses.
Fig 7.
Trends observed in the estimated measurements (CI, GI, head volume, and height/half-length (h/a2)) at 3 months over the period from 2010 to 2019.
(a) CI, (b) GI, (c) head volume, and (d) h/a2. A letter was allocated to each birth year from left to right as indicated inside the box. The letter assigned to the group that came first in the sequence was added above the boxplot of the group that appear later in the sequence if the difference between estimated marginal means was significant at either p < 0.05 (in black), p < 0.01 (in blue), or p < 0.001 (in red) level based on the pairwise post hoc t-test with FDR correction. 2010 (n = 52), 2011 (n = 100), 2012 (n = 148), 2013 (n = 25), 2014 (n = 52), 2015 (n = 15), 2016 (n = 19), 2017 (n = 51), 2018 (n = 63), and 2019 (n = 48). See S33-S44 Tables in S1 File for details of post hoc analyses.
Fig 8.
Comparative review of head circumference (HC) measurements from this study and the national growth survey on preschool children for 1960 to 2010.
A scatter plot of all HC measurements (dots in green) including infants whose age did not correspond to 2, 3, 6 or 12 months but removing infants older than 375 days to match the age range of the national growth survey data (n = 1,958). A logarithmic approximation curve (a curve in black) was overlayed with government data (lines in colors) made available for the period from 1960 to 2010. As the national growth survey data only provided the average measurement for the corresponding age range and sex, the mean values combining both sexes were plotted against the midpoints of the age range.