Table 4 Summary of the fetal imaging data and cohort demographics (covering a range from a minimum of 18.0 to a maximum of 36.8 gestational weeks) used in this study.

Measure	All data	Male scans	Female scans	T statistic P value	Effect size
Num. of scans	268	139	129	–	–
Num. of fetuses	162	81	81	–	–
Num. of fetuses with one scan	88	41	47	–	–
Num. of fetuses with two scans	42	22	20	–	–
Num. of fetuses with three scans	32	18	14	–	–
Gestational age at scan (weeks)	27.63 (±5.12)	27.53 (±5.12)	27.74 (±5.14)	0.7456	0.040
Maternal age (years)	33.24 (±4.21)	33.25 (±3.69)	33.22 (±4.75)	0.9572	0.007
Paternal age (years)	35.50 (±5.96)	35.64 (±5.48)	35.34 (±6.45)	0.7019	0.049
Maternal height (m)	5.44 (±0.26)	5.46 (±0.26)	5.42 (±0.27)	0.1712	0.169
Maternal weight before (kg)	65.99 (±13.34)	66.67 (±14.50)	65.27 (±12.00)	0.4056	0.104
Household income (USD)	175 K (±124 K)	152 K (±92K)	200 K (±148 K)	0.0052	0.393
Num. in household	2.58 (±0.92)	2.42 (±0.74)	2.76 (±1.07)	0.0044	0.365
Num. of previous pregnancies	1.19 (±1.52)	1.07 (±1.34)	1.30 (±1.70)	0.2250	0.151
Num. of previous miscarriages	0.50 (±0.82)	0.46 (±0.84)	0.55 (±0.80)	0.3778	0.111
Num. of previous abortions	0.13 (±0.49)	0.13 (±0.57)	0.13 (±0.40)	0.9512	0.008
Previous ectopic pregnancies	0.02 (±0.15)	0.00 (±0.00)	0.03 (±0.22)	0.1027	0.211
Num. of previous still births	0.02 (±0.16)	0.02 (±0.15)	0.02 (±0.18)	0.7498	0.040

The overall and separate average (and standard deviation) for male and female cohorts are included for the contributing scans in each group. These are based on the subset of cases where the parents chose to provide answers. The only measures that exhibited statistically a two-sided difference (P < 0.05, uncorrected significance in bold) in means between the male and female scans that could potentially alternatively explain male vs. female differences in brain anatomy were the household income and number in household. Both of these were statistically greater for the female fetuses, but with only a small to medium effect size (Cohen’s D). We therefore explored the incorporation of these measures as covariates in the statistical analysis, and found that this did not significantly impact the findings. However, as not all participants provided responses to these questions, this would reduce the number of imaging measures; therefore, in the full cohort analysis we do not include these covariates.

Search