Three hundred households with a total of 1056 household members and a median household size of 4 [IQR 3-5] household members were enrolled of which 991 (94%) participated in this study. Median age of the participants was 33 [IQR 15-46], 341/991 (34.4%) were <18 years of age. In total, 950 participants consented to blood collection and were tested for antibodies against SARS-CoV-2 of which 587 (61.8%) were seropositive. An index case was defined as the person that was most probable the first one infected according to onset of symptoms, positive test or seropositivity and was identified in 276/300 (92%) households. Thereof, 244 (88.4%) reported a symptomatic infection with a significant difference between index cases <18 years old and index cases ≥18 years old (55.6% vs 94.8%; p<.0001). The ratio of index cases <18 (16.3%) and ≥18 years old (83.7%) was also clearly different. 229/276 (83.0%) index cases were seropositive in our analysis. This ratio was the same for index cases <18 and ≥18 years of age. 370 (63%) of all seropositive participants reported a previous positive SARS-CoV-2 PCR. In addition, participants with comorbidities were more likely to be seropositive (68.3% vs 55.9%, p<.001) and to have symptomatic SARS-CoV-2-infections (69.4% vs 56.4%; p<.0001) compared to those without. Further demographics of the study population are listed in Table 1.
Table 1
Descriptive Statistics of the Study Population
|
Households, total
|
300
|
|
Household size, median (IQR)
|
4
|
(3-5)
|
Household members, total
|
1056
|
|
Study Participants
|
991
|
|
|
≥ 18 years
|
650
|
(66%)
|
|
< 18 years
|
341
|
(34%)
|
Age, median (IQR)
|
33
|
(15-46)
|
|
≥ 18 years
|
43
|
(34-49)
|
|
< 18 years
|
11
|
(7-15)
|
Sex, male
|
482
|
(49%)
|
Households with Index Case
|
276
|
|
|
Index ≥ 18 years
|
231
|
(84%)
|
|
|
seropositive
|
194
|
(84%)
|
|
|
symptomatic
|
219
|
(95%)
|
|
Index < 18 years
|
45
|
(16%)
|
|
|
seropositive
|
35
|
(78%)
|
|
|
symptomatic
|
25
|
(56%)
|
Symptomatic COVID-19
|
|
|
|
Participants with comorbidities
|
197
|
(69%)
|
|
Participants without comorbidities
|
390
|
(56%)
|
Regarding transmissions within the households, in 27/299 (9%) households none of the participants were seropositive and in 58/299 (19.4%) households only the index-case was seropositive. In 112/299 (37.5%) households all household members were seropositive. There was a lower chance that everyone in the household was seropositive with a minor index case than with an index case ≥18 years (18.2% vs 43.8%; p=.0013) (Table 2a). Accordingly, the SAR of underage index cases was significantly lower compared to ≥18 years old index cases (0.29 vs 0.54; p<.0001) (Table 3).
Table 2
Household contagion after SARS-CoV-2-infection
a. to different household characteristics
Households [%]
|
all households
|
with <18 y/o
|
without <18 y/o
|
Index <18 y/o
|
Index ≥18 y/o
|
Symptomatic Index
|
Asymptomatic Index
|
2 members
|
3 members
|
4 members
|
>4 members
|
everyone seropositive
|
37.6
|
22.2
|
61
|
18.2
|
43.8
|
43.9
|
0
|
63.9
|
36.4
|
23.3
|
17.6
|
some seropositive
|
34.2
|
47.7
|
14.4
|
29.5
|
30.5
|
33.5
|
10.5
|
5.6
|
36.4
|
50
|
70.6
|
only index seropositive
|
19.7
|
21.6
|
16.1
|
38.6
|
17.7
|
16.7
|
42.1
|
18.1
|
20
|
20
|
5.9
|
no one seropositive
|
8.5
|
8.5
|
8.5
|
13.6
|
8
|
5.9
|
47.4
|
12.5
|
7.27
|
6.7
|
5.9
|
|
|
χ² (3, 294) = 52.18; p<.00001
|
χ² (3, 270) = 15.09; p=.0017
|
χ² (3, 258 = 51.37; p<.00001
|
χ² (9, 231) = 58.09; p<.00001
|
b. according to hygiene measures
Households [%]
|
no hygiene measures
|
unaware of infection
|
any hygiene measures
|
Index moved out
|
seperate rooms
|
seperated use of rooms
|
physical distance
|
hand hygiene
|
air ventilation
|
face mask
|
everyone seropositive
|
43.4
|
7.5
|
42.5
|
16.7
|
27
|
40.7
|
42.9
|
46.7
|
46.1
|
30
|
some seropositive
|
33.2
|
45
|
27.5
|
33.3
|
37.8
|
33.3
|
34.8
|
31.5
|
31.8
|
32
|
only index seropositive
|
16.1
|
40
|
15
|
33.3
|
24.3
|
16.7
|
13.4
|
15.8
|
14.3
|
24
|
noone seropositive
|
7.3
|
7.5
|
15
|
16.7
|
10.8
|
7.4
|
6.3
|
6.1
|
7.8
|
10
|
|
|
χ²(3,80)=17.04
p<.001
|
χ²(3,251)=2.67 p=.45
|
χ²(3,46)=1.97 p=.58
|
χ²(3,77)=3.06 p=.38
|
χ²(3,94)=1.54 p=.67
|
χ²(3,152)=3.11 p=.37
|
χ²(3,209)=3.60 p=.31
|
χ²(3,199)=2.07 p=.56
|
χ²(3,90)=2.43 p=.49
|
Note: Chi-Square was calculated compared to households with no hygiene measures.
Table 3
SARS-CoV-2 secondary attack rates (SAR)
|
overall
|
0.53
|
Index case
|
|
|
<12 years
|
0.38
|
|
12-18 years
|
0.26
|
|
<18 years
|
0.29 *
|
|
18-40 years
|
0.44
|
|
>40 years
|
0.60
|
|
≥18 years
|
0.54
|
Index <18 years
|
|
|
contacts <18 years
|
0.29
|
|
contacts ≥18 years
|
0.29 Ⴕ
|
Index ≥18 years
|
|
|
contacts <18 years
|
0.48
|
|
contacts ≥18 years
|
0.59
|
Index case
|
|
|
symptomatic
|
0.58
|
|
asymptomatic
|
0.02 ٭
|
Household
|
|
|
aware of infection
|
0.52
|
|
unaware of infection
|
0.22 ˣ
|
* p<.0001 compared to index cases ≥18 years
Ⴕ p=.017 compared to index ≥18 years to contacts <18 years
٭ p<.0001 compared to symptomatic index cases
ˣ p<.0001 compared to households aware of infection
Specifically, transmission from adult index cases to household contacts <18 years was more common than vice versa (SAR 0.48 vs 0.29; p=.017). Children and adolescents generally transmitted SARS-CoV-2-infections less frequently to both adult and underage household contacts (SAR 0.29). When comparing transmissions within households with PCR confirmed SARS-CoV-2-infections to those in which the SARS-CoV-2-infection was detected only retrospectively by serology, we found that SAR was significantly higher within households with a positive PCR (0.52 vs 0.22; p<.0001). Even when a household did not follow any hygiene measures during quarantine, SAR was higher than in households that were unaware of their SARS-CoV-2-infection (0.53 vs 0.22; p<.0001). Accordingly, in households with a symptomatic index person there was a higher chance that every participant from the household was seropositive (44.1% vs 0%; p<.0001). The seropositivity rate was also dependant of the household size as it decreased significantly for each additional household member. In households with more than 4 members the probability of all household members being seropositive was significantly lower compared to households with only two members (17.6% vs. 63.9%; p<.0001). To analyze the effect of hygiene measures on transmission, participants were asked if they were following specific hygiene measures e.g. distancing, wearing masks or regular room ventilation. Interestingly, the overall SAR was the same in households that implemented hygiene measures to those that did not (0.53) (Table 2b). When the index case left the household (6/211 households) as soon as the infection was confirmed, there was a 33.3% chance that no other household member was seropositive (SAR of 0.36), while this was only the case in 15% of the households without any measures (SAR 0.53). Due to the low case numbers, this was not statistically significant.
To evaluate the severity and long-term effect of the infection, participants were asked about current medical conditions at the time of blood collection. Results are displayed in Table 4 as ongoing symptoms <12 weeks and ≥12 weeks after SARS-CoV-2 infection according to specific characteristics of the participants. When the infection was <12 weeks ago 25 % of the participants suffered from ongoing symptoms while participants that were enrolled ≥12 weeks after their infection reported health restrictions in 29.7%. Seropositive participants were more likely to have persistent symptoms both <12 and ≥12 weeks after infection within their household compared to seronegative participants (12 weeks: 31.5% vs 14.5% and ≥12 weeks: 34.8% vs 20.8%; p<.001). When analysing only participants <18 years of age, differences between seronegative and seropositive participants were not significant. Individuals that reported SARS-CoV-2 related symptoms in the acute phase were more often suffering from ongoing symptoms than asymptomatic participants both <12 and ≥12 weeks after the event of SARS-CoV-2-infection within the household (< 12 weeks: 33.3% vs 5.0% and >12 weeks: 39.8% vs 3.1%; p<.0001).
Table 4
Ongoing symptoms after SARS-CoV-2-infection
time after SARS-CoV-2
|
ongoing symptoms
|
male
|
female
|
<18 years
|
≥18 years
|
<12 years
|
12 - 17 years
|
18 - 40 years
|
>40 years
|
<12 weeks
|
yes
|
19.7
|
30.1
|
8.2
|
32.6
|
6.3
|
10.8
|
21.3
|
40.6
|
no
|
80.3
|
69.9
|
91.8
|
67.4
|
93.8
|
89.2
|
78.7
|
59.4
|
|
|
p<.05
|
p<.0001
|
ns
|
p<.01
|
≥12 weeks
|
yes
|
17.5
|
40.4
|
3.8
|
41.9
|
0
|
7.3
|
37.8
|
45.5
|
no
|
82.5
|
59.6
|
96.2
|
58.1
|
100
|
92.7
|
62.2
|
54.5
|
|
|
p<.0001
|
p<.0001
|
p<.01
|
p=.053
|
|
|
|
|
|
|
|
|
|
|
time after SARS-CoV-2
|
ongoing symptoms
|
seropositive
|
seronegative
|
symptomatic COVID-19
|
asymptomatic COVID-19
|
comorbidities
|
no comorbidities
|
BMI < 25
|
BMI > 25 *
|
<12 weeks
|
yes
|
31.5
|
14.5
|
33.3
|
5
|
42.5
|
17.8
|
8.2
|
0
|
no
|
68.5
|
85.5
|
66.7
|
95
|
57.5
|
82.2
|
91.8
|
100
|
|
|
p<.001
|
p<.0001
|
p<.0001
|
ns
|
≥12 weeks
|
yes
|
34.8
|
20.8
|
39.8
|
3.1
|
41
|
24.5
|
2.2
|
22.2
|
no
|
65.2
|
79.2
|
60.2
|
96.9
|
59
|
75.5
|
97.8
|
77.8
|
|
|
p<.001
|
p<.0001
|
p<.01
|
p<.05
|
Note: table shows fraction of participants with and without ongoing symptoms [%], p-values refer to Fisher’s exact test results, * BMI analysis for participants <18 years old, n=178
Generally, adults were more likely to develop ongoing symptoms compared to children and teenagers <18 years (<12 weeks: 32.6% vs 8.2%; p<.0001 and ≥12 weeks: 41.9% vs 3.8%; p<.0001). Furthermore, participants >40 years of age were more likely to have ongoing symptoms up to 12 weeks after infection than young adults (46.9% vs 21.3%; p<.01). When participants ≥18 years were enrolled ≥12 weeks after the infection, they more often reported medical issues than those seen within 12 weeks after infection (41.9% vs 32.6%) while the rate of persisting symptoms in minor participants decreased over time (<12 weeks 8.2% vs >12weeks 3.8%). The probability of ongoing symptoms did not only correlate with age but with gender. Female participants had a higher prevalence of ongoing symptoms ≥12 weeks after the infection compared to male participants (40.4 % vs 17.5 %; p<.0001). Consistent with higher infection rates and seroprevalence, participants with comorbidities were more likely to experience persisting symptoms after the infection (< 12 weeks: 42.5% vs 17.8%; p<.0001 and >12 weeks: 41.0% vs 24.5%; p<.01). Here, seropositive participants were significantly more affected than seronegative subjects (47.5% vs 21.6%; p<.005). BMI did not have any influence on the rate of ongoing symptoms in participants ≥18 years old, but children and adolescents with a BMI >25 were more likely to report ongoing symptoms ≥12 weeks after the infection than those of lower BMI (22,2% vs 2.2%, p<.05).