Table 3.

Final multivariable logistic regression model showing the association between potential determinants and the presence of Methanosphaera stadtmanae.

M. stadtmanae presence
Adjusted main effects model (N = 420)a
Determinant	OR (CI95%)	FDR crit.b	p-value
Birthweight (g)c	1.001 (1.000; 1.002)	0.019	0.028
Place and mode of deliveryc
Natural birth at home	Ref.
Natural/artificial birth at hospital	1.61 (0.67; 3.88)	0.034	0.288
Caesarean section	6.89 (2.09; 22.67)	0.003	0.002f
First exposure to antibiotics (during first 2 years of life)c
Never	Ref.
At 0–7 months	0.54 (0.16; 1.83)	0.038	0.323
At 8–12 months	0.70 (0.21; 2.36)	0.050	0.566
At 13–24 months	3.38 (1.34; 8.50)	0.013	0.010f
Regular intake of organic milkd,e
No	Ref.
Yes	3.52 (0.97; 12.73)	0.022	0.056
Regular intake of organic yogurtd,e
No	Ref.
Yes	1.690 (0.49; 5.86)	–	0.409
Regular intake of organic fruitc
No	Ref.
Yes	4.73 (1.64; 13.62)	0.006	0.004f
Animal protein intake (en%)c	0.77 (0.65; 0.92)	0.009	0.004f

^aModel adjusted for: age at fecal sampling (years), gender (male/female), recruitment group (conventional/alternative), total energy intake (kcal) and BMI (z-score).

^bCritical FDR cut-off level as determined by Benjamini-Hochberg procedure.

^cParameter estimates presented for the model containing organic milk intake. Parameter estimates did not change substantially when the variable “organic milk intake” was replaced by the variable “organic yogurt intake” (data not shown).

^dMissing value category omitted (included in FDR correction).

^eDue to multicollinearity both regular intake of organic milk and regular intake of organic yogurt were included in separate models, which included all other variables as listed in this table.

^fSignificant association after correction for FDR by Benjamini-Hochberg procedure, also indicated in bold.