Table 2.

Logistic regression analysis assessing the association of living with children and SARS-CoV-2 seropositivity

Exposure	Odds ratio (95% CI)
Model 0: Any child in household, unadjusted analysis (2375 complete cases)
No child in household	[Reference]
≥ 1 child in household	1.22 (0.98–1.52)
Model 1: Any child in household, adjusted analysis (2232 complete cases)
No child in household	[Reference]
≥ 1 child in household	1.25 (0.99–1.58)
Model 2: Age of children in household (2232 complete cases, 2114 included a))
No children in household	[Reference]
Only children 0–11 years	1.21 (0.90–1·60)
Only children 12–17 years	1.14 (0.78–1·64)
Model 3: Number of children in household (2232 complete cases)
Number of children in household (per child)	1.14 (1.02–1.27)
Model 4: Interaction of living with children and out of household contacts (2197 complete cases)
No child in household	[Reference]
≥ 1 child in household	1.21 (0.69–2.10)
No non-household contact in previous 7 days	[Reference]
≥ 1 non-household contact in previous 7 days	0.97 (0.66–1.47)
Interaction	1.06 (0.58–1.96)
Model 5: Interaction of living with children and employment status (2122 complete cases)
No child in household	[Reference]
≥ 1 child in household	1.44 (1.06–1.93)
Working outside home	[Reference]
Being unemployed or working full-time from home	0.69 (0.51–0.94)
Interaction	0.67 (0.41–1.09)
Model 6: Interaction of living with children and gender of participant (2232 complete cases)
No child in household	[Reference]
≥ 1 child in household	0.95 (0.68–1.32)
Female gender	[Reference]
Male gender	0.87 (0.65–1.16)
Interaction	1.74 (1.10–2.76)

Models 1–6 are adjusted for age, gender, income, education, and smoking

^a88 cases were excluded as they included children of both age groups (living with children 0–11 years and 12–17 years)