Sex Differences in the Association of Sibship Size and Position in Sibship with Lipid Profile during Adolescence: A Cross-Sectional Study

Table 4

Geometric means and adjusted ratio of geometric means of high-density lipoprotein cholesterol according to the total number of siblings, number of older siblings, and number of younger siblings stratified by sex.

	Outcome variable: High-density lipoprotein cholesterol						Sex-sibling ^#
	Boys			Girls
	n	Geometric mean (mmol/L)	Ratio of geometric means (95% CI)^‡	n	Geometric mean (mmol/L)	Ratio of geometric means (95% CI)^‡
Total siblings							0.203
0 and 1^†	21	1.08	1.00 (reference)	22	1.32	1.00 (reference)
2	52	1.23	1.14 (0.99–1.30)	44	1.29	0.98 (0.87–1.10)
3	119	1.14	1.06 (0.93–1.20)	98	1.33	1.01 (0.91–1.13)
≥4	527	1.11	1.02 (0.91–1.15)	644	1.27	0.96 (0.87–1.06)
Older siblings							0.625
0	187	1.12	1.00 (reference)	192	1.27	1.00 (reference)
1	146	1.10	0.98 (0.92–1.04)	158	1.25	0.99 (0.94–1.04)
2	139	1.13	1.01 (0.95–1.07)	154	1.26	1.00 (0.95–1.05)
≥3	277	1.09	0.97 (0.92–1.03)	316	1.29	1.02 (0.97–1.07)
Younger siblings							0.094
0	57	1.04	1.00 (reference)	58	1.29	1.00 (reference)
1	118	1.12	1.07 (0.99–1.17)	145	1.28	1.00 (0.93–1.07)
2	144	1.14	1.10 (1.02–1.19)^§	138	1.24	0.97 (0.90–1.04)
≥3	397	1.13	1.08 (1.01–1.17)^§	461	1.25	0.98 (0.91–1.04)

CI: confidence interval. ^†Since only one child had zero siblings, we combined the no siblings and one sibling groups: reported as: 0 and 1. ^‡Adjusted for age, body mass index-for-age, mode of birth, breastfeeding status, smoking status, frequency of vigorous physical activity per week, exposure to household secondhand smoke, prediabetes status, parental history of diabetes, maternal education level, and paternal education level. Additionally, ratios of geometric means of older siblings were simultaneously adjusted for younger siblings, and ratios of geometric means of younger siblings were simultaneously adjusted for older siblings. ^§-value <0.05. ^#-value evaluating statistical interaction on multiplicative scale between sex and number of total, older, and younger siblings in the total study sample. An interaction -value <0.2 was considered to be a “possible” statistical suggestion for interaction.