Frequency of human enterovirus fecal samples collected before islet autoimmunity
| . | Case subjects . | Control subjects . | OR (95% CI)* . | |
|---|---|---|---|---|
| Unadjusted . | Adjusted† . | |||
| n | 27 | 53 | ||
| Enterovirus RNA | ||||
| Negative samples | 296 | 598 | 1.00 (reference) | 1.00 (reference) |
| Positive samples | 43 (12.7) | 94 (13.6) | 1.01 (0.59–1.72) | 1.09 (0.61–1.96) |
| Total | 339 | 692 | ||
| New enterovirus infection episode | ||||
| No | 296 | 598 | 1.00 (reference) | 1.00 (reference) |
| Yes | 30 (9.2) | 65 (9.8) | 0.94 (0.59–1.52) | 0.92 (0.54–1.57) |
| Total‡ | 326 | 663 | ||
| . | Case subjects . | Control subjects . | OR (95% CI)* . | |
|---|---|---|---|---|
| Unadjusted . | Adjusted† . | |||
| n | 27 | 53 | ||
| Enterovirus RNA | ||||
| Negative samples | 296 | 598 | 1.00 (reference) | 1.00 (reference) |
| Positive samples | 43 (12.7) | 94 (13.6) | 1.01 (0.59–1.72) | 1.09 (0.61–1.96) |
| Total | 339 | 692 | ||
| New enterovirus infection episode | ||||
| No | 296 | 598 | 1.00 (reference) | 1.00 (reference) |
| Yes | 30 (9.2) | 65 (9.8) | 0.94 (0.59–1.52) | 0.92 (0.54–1.57) |
| Total‡ | 326 | 663 | ||
Data are n, n (%), and ORs (95% CI).
*Estimated from logistic mixed-effects logistic regression models with random intercept for each subject to control for intraindividual correlation (no significant random intercept in model for enterovirus episodes, but highly significant in model for enterovirus positivity). The unadjusted OR in ordinary logistic regression ignoring intraindividual correlation in infections was 0.92;
‡Adjusted for sex, calendar month of sample collection, year of sample collection (2001–2003, 2004–2006, or 2007–2008), age (continuous), number of siblings (0 vs. ≥1), breast-feeding, and first-degree family history of type 1 diabetes (yes/no);
†Excluding consecutively positive samples that may have been part of the same infectious episode as in the previous positive sample.