TABLE 2.

Progression from islet autoimmunity to clinical type 1 diabetes in sample interval (median ∼4 months) following infection detected by enterovirus RNA in serum or rectal swab sample

Type of sample	Person-years of follow-up	Cases progressing to type 1 diabetes in interval*	Unadjusted HR (95% CI)	HR (95% CI) adjusted for islet autoantibodies
Serum
No enterovirus RNA in previous sample	494	33	1.00 (ref.)	1.00 (ref.)
Enterovirus RNA in previous sample	6.5	3	6.36 (1.89–21.4)†	7.02 (1.95–25.3)
Rectal swab
No enterovirus RNA in previous sample	537.1	32	1.00 (ref.)	1.00 (ref.)
Enterovirus RNA in previous sample	21.2	1	0.93 (0.12–6.90)	0.79 (0.10–5.92)

Data are n unless otherwise indicated.

*Forty-one of 140 children in the study cohort progressed to type 1 diabetes during the period wherein collected samples were tested for enterovirus, of which serum enterovirus RNA results were available from the clinic visit preceding diagnosis in 36 (of which 3 were positive) and rectal swab enterovirus RNA was available in 33 (of which 1 was positive, and serum from the same visit was also positive for enterovirus RNA). Enterovirus exposure variables coded according to the rapid effect model described in research design and methods.

†Cox regression model: P = 0.003. Permutation test based on Cox regression model with 10,000 permutations of the enterovirus variable: P = 0.0075.