Examine age differences in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission risk from primary cases and infection risk among household contacts and symptoms among those with SARS-CoV-2 infection.
People with SARS-CoV-2 infection in Nashville, Tennessee and central and western Wisconsin and their household contacts were followed daily for 14 days to ascertain symptoms and secondary transmission events. Households were enrolled between April 2020 and April 2021. Secondary infection risks (SIR) by age of the primary case and contacts were estimated using generalized estimating equations.
The 226 primary cases were followed by 198 (49%) secondary SARS-CoV-2 infections among 404 household contacts. Age group-specific SIR among contacts ranged from 36% to 53%, with no differences by age. SIR was lower in primary cases age 12 to 17 years than from primary cases 18 to 49 years (risk ratio [RR] 0.42; 95% confidence interval [CI] 0.19–0.91). SIR was 55% and 45%, respectively, among primary case-contact pairs in the same versus different age group (RR 1.47; 95% CI 0.98–2.22). SIR was highest among primary case-contact pairs age ≥65 years (76%) and 5 to 11 years (69%). Among secondary SARS-CoV-2 infections, 19% were asymptomatic; there was no difference in the frequency of asymptomatic infections by age group.
Both children and adults can transmit and are susceptible to SARS-CoV-2 infection. SIR did not vary by age, but further research is needed to understand age-related differences in probability of transmission from primary cases by age.
Considerable heterogeneity regarding reports on transmission from and susceptibility of children to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) relative to adults. Clinical data from nonhospitalized children are limited, with few studies reporting symptoms and testing systematically and frequently.
Both children and adults can transmit and are susceptible to SARS-CoV-2 infection. Susceptibility to SARS-CoV-2 was high and similar across all age groups in the household setting. There was no difference in frequency of asymptomatic infections between children and adults.
Whereas some studies suggest children are less susceptible to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection than adults,1–4 other studies of household or other close contacts have found similar secondary SARS-CoV-2 infection rates among children and adults.5,6 Variation in mixing patterns and likelihood of exposure and detection (contact tracing and testing practices) may contribute to reported differences in children versus adults.2,7 Most previous reports on SARS-CoV-2 infection in children used surveillance or contact tracing data and were conducted early in the course of the pandemic. During that time, children were largely protected from community exposures due to nonpharmaceutical interventions, including closure of businesses and schools. Furthermore, many of those early studies were conducted in Asia, where prevention efforts and age-related interactions likely differed from those in the United States.1–3
Children tend to have less severe illness than adults,8–11 but the spectrum of illness and SARS-CoV-2 transmission risk have not been fully characterized among children in US households. Additionally, coronavirus disease 2019 (COVID-19) cases and COVID-19-associated hospitalizations among children have increased significantly as the Delta variant became the predominant circulating strain.12,13 Greater understanding of age-related differences in susceptibility, transmission risk, and illness characteristics, particularly in children, is needed to guide public health recommendations on prevention of transmission and inform plans for in-person school attendance.
We previously reported on SARS-CoV-2 transmission from a prospective study of US households.14 We extended those results and examined age differences in SARS-CoV-2 transmission risk from primary cases and infection risk among household contacts. We also assessed age-specific differences in symptoms and illness duration among secondary cases.
Methods
Design, Setting, and Participants
This analysis used data from a prospective case-ascertained household SARS-CoV-2 transmission study conducted in Tennessee and Wisconsin between April 21, 2020 and April 30, 2021.14,15 Persons with laboratory-confirmed SARS-CoV-2 infection (index cases) were identified from SARS-CoV-2 clinical real-time reverse transcription polymerase chain reaction (rRT-PCR) tests conducted at Vanderbilt University Medical Center (VUMC; Nashville, Tennessee) and Marshfield Clinic Health System (MCHS; Marshfield, Wisconsin). VUMC is a large health care provider system serving patients from Tennessee and the Mid-South US. MCHS is a large community-based, multispecialty health care system serving predominantly rural populations in central, northern, and western Wisconsin. For this study, we enrolled patients presenting to VUMC walk-in-clinics that operate within Davidson County and surrounding areas and MCHS locations in central and western Wisconsin. Most schools in the study area were closed to in-person attendance in Spring 2020. In Fall 2020, some schools (public and private in Wisconsin and private in Tennessee) were open for full-time in-person attendance or had hybrid modalities (combination of in-person or remote attendance). SARS-CoV-2 testing capacity at both sites varied throughout the study period. Testing was limited in Spring 2020, increased by Summer 2020, with return times for results taking longer during periods with a high-level of community transmission. By Fall 2020, SARS-CoV-2 testing services were readily available with rapid return of results.16
Index cases (ie, the first person in the household identified with a positive rRT-PCR SARS-CoV-2 result) and their household contacts were followed daily for 14 days to ascertain symptoms and secondary transmission events. Households were eligible if the index case had symptom onset <7 days before enrollment and there was ≥1 other household member without symptoms at the time of the index case’s illness onset. The primary case was the person with laboratory-confirmed SARS-CoV-2 infection in the household with the earliest illness onset date (or date of positive SARS-CoV-2 sample, if asymptomatic). Coprimary cases were household members positive for SARS-CoV-2 with illness onset occurring or first positive SARS-CoV-2 sample collected within 2 days after illness onset in the primary case.
Data and Sample Collection
Data were primarily collected through self- or parent-administered paper (Wisconsin) or web-based surveys (Tennessee) (Supplemental Information); some data were obtained through interviews with participants. At enrollment, the survey assessed demographic and household characteristics, preexisting medical conditions, occupational risk (eg, employed in health care setting or customer service), symptoms before enrollment, and type and frequency of interactions with other household members. During the 14-day follow-up period, participants were asked about current symptoms and provided a (self- or parent-collected) respiratory (anterior nasal) and/or saliva sample for SARS-CoV-2 rRT-PCR testing every day regardless of symptoms. Symptoms assessed included constitutional (chills, fatigue or feeling run down, fever or feverishness, muscle or body aches), upper respiratory (nasal congestion, runny nose, sore throat), lower respiratory (chest tightness or pain, cough, trouble breathing or shortness of breath, wheezing), neurologic (headache, loss of taste or smell), and gastrointestinal (abdominal pain, diarrhea, vomiting). All participants were asked about all symptoms except gastrointestinal, which was only included in Wisconsin.
Laboratory
Analysis
Participants were grouped by age to reflect potential exposure risk and behavioral characteristics: preschool-aged (0–4 years), primary school–aged (5–11 years), secondary school–aged (12–17 years), young adults (18–49 years), middle-aged adults (50–64 years), and older adults (≥65 years).
Differences among age groups were assessed using χ2 test or Kruskal-Wallis test, where appropriate. Secondary SARS-CoV-2 infections were defined as household contacts with ≥1 rRT-PCR positive sample (respiratory or saliva) with the illness onset or first positive sample date occurring within 14 days after the illness onset (or date of first positive sample, if asymptomatic) in the primary case. We estimated secondary infection risks (SIR) (1) by age of the primary case to assess transmission risk (the probability of transmission from the primary case to contacts), (2) by age of the contacts to assess infection risk (the probability of infection among contacts), (3) by age of the primary case and age of the contacts, and (4) by whether the primary case and contact were in the same or different age group. SIRs were estimated using a generalized estimating equation (GEE) log-binomial model that included covariates for age of the primary case, age of the contacts, and whether the primary case and contact were in the same age group, accounting for household clustering. Risk ratios (RR) and 95% confidence intervals (CI) were used to compare SIR in each primary and contact age group versus age 18 to 49 years (referent) and same versus different age groups. Contacts who had received COVID-19 vaccine before enrollment, had an illness onset >10 days before enrollment (as infections may have become undetectable), had an illness onset before the primary case without laboratory-confirmed SARS-CoV-2 infection, or had <6 follow-up days with survey data or <6 follow-up days with samples with rRT-PCR results were excluded. Households with ≥1 coprimary cases, vaccinated primary case, or had no remaining eligible household contacts were excluded.
We assessed frequency, sequence, and duration of symptoms by age group among participants with SARS-CoV-2 infection whose illness onset or first positive sample (if asymptomatic) occurred after study enrollment. Associations between age group and presence of specific categories of symptoms were assessed using logistic regression models. Odds ratios (OR) and 95% CI were used to compare odds of symptoms in each age group versus age 18 to 49 years (referent). Neurologic symptoms were not assessed in children age 0 to 4 years, as these symptoms are difficult to ascertain in young children. Analysis of symptom duration and interval (in days) between symptom onset and first positive sample was restricted to symptomatic infections. Analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC).
The study protocol was approved by institutional review boards at VUMC and MCHS. CDC determined this activity was conducted consistent with applicable federal law and CDC policy (see 45 C.F.R. part 46; 21 C.F.R. part 56).
Results
Participant Characteristics
From April 2020 through April 2021, 302 index cases and their 577 household members were enrolled. The index case was the primary case for 96% of households. The primary analysis included 226 of 302 (75%) households with 404 contacts. Reasons for exclusion are shown in Fig 1.
The majority of primary cases were non-Hispanic White young adults with a preponderance of female participants (Table 1). The median age was 37 years (range: 1–76 years). An underlying medical condition was reported by 81 (36%), of whom 29 (36%) had asthma. Among adults, 20% reported working in a health care setting and had regular face-to-face contact with sick people and 19% in customer service with regular face-to-face contact with people. Most primary cases (83%) lived in a single-family home with mean of 3.2 bedrooms and 3.3 members. Among those who reported on interactions with other household members, interactions decreased from the day before illness onset to the day before enrollment; 73% reported physical contact with ≥1 other household member the day before illness onset versus 38% the day before enrollment (Table 1). Masking when interacting with other members was uncommon before illness onset (6%), and 26% reported mask use the day before enrollment.
. | Age, y . | ||||||
---|---|---|---|---|---|---|---|
0–4 . | 5–11 . | 12–17 . | 18–49 . | 50–64 . | ≥65 . | All Ages . | |
N (row %) | 2 (0.9) | 4 (1.8) | 25 (11.1) | 141 (62.4) | 37 (16.4) | 17 (7.5) | 226 (100) |
Female | 2 (100) | 0 (0) | 15 (60) | 81 (57) | 21 (57) | 9 (53) | 128 (56.6) |
Race/ethnicitya | |||||||
Non-Hispanic White | b | b | 23 (92) | 115 (82) | 35 (95) | 17 (100) | 194 (85.8) |
Non-Hispanic other race | b | b | 0 | 9 (6) | 1 (3) | 0 | 11 (4.9) |
Hispanic | b | b | 2 (8) | 16 (11) | 1 (3) | 0 | 20 (8.9) |
Smoker (age ≥18 y) | — | — | — | 7 (5) | 2 (5) | 0 | 9 (5) |
Any underlying medical conditionsc | 1 (50) | 0 | 9 (36) | 46 (33) | 13 (35) | 12 (71) | 81 (35.8) |
Occupational, school, or child care exposures | |||||||
Attended child care or school outside homed (age <18 y) | 1 (100) | 3 (100) | 14 (74) | — | — | — | 18 (78) |
Health care setting (age ≥18 y)d,e | — | — | — | 33 (24) | 6 (16) | 0 | 39 (20) |
Customer service (age ≥18 y)d,e | — | — | — | 31 (22) | 4 (11) | 2 (12) | 37 (19) |
Teacher (age ≥18 y)d | — | — | — | 4 (4) | 2 (7) | 1 (8) | 7 (5) |
Household characteristics | |||||||
Mean (SD) number of household membersf | 4.0 (0) | 4.3 (1.7) | 4.4 (0.9) | 3.4 (1.4) | 2.7 (0.8) | 2.1 (0.2) | 3.3 (1.4) |
Type of home | |||||||
Single family home | 2 (100) | 3 (75) | 25 (100) | 110 (78) | 33 (89) | 14 (82) | 187 (82.7) |
Duplex/townhome | 0 | 0 | 0 | 0 | 0 | 8 (3.5) | |
Condominium/apartment building | 0 | 1 (25) | 0 | 23 (16) | 4 (11) | 3 (18) | 31 (13.7) |
Mean (SD) number of bedroomsf | 4.0 (0) | 3.3 (1.0) | 3.9 (0.7) | 3.0 (1.0) | 3.4 (1.2) | 2.9 (0.7) | 3.2 (1.0) |
Interactions with other household membersd,g | |||||||
Maximum time spent in same room with ≥1 other member | |||||||
Day before illness onset | |||||||
>4 h | 2 (100) | 2 (100) | 10 (43) | 56 (59) | 10 (40) | 6 (60) | 86 (55) |
1–4 h | 0 | 0 | 7 (30) | 23 (24) | 11 (44) | 1 (10) | 42 (27) |
<1 h | 0 | 0 | 3 (13) | 16 (17) | 4 (16) | 3 (30) | 26 (17) |
No time | 0 | 0 | 3 (13) | 0 | 0 | 0 | 3 (2) |
Day before enrollment | |||||||
>4 h | 2 (100) | 1 (50) | 4 (17) | 35 (37) | 6 (23) | 2 (20) | 50 (32) |
1–4 h | 0 | 0 | 7 (30) | 15 (16) | 6 (23) | 3 (30) | 31 (20) |
<1 h | 0 | 1 (50) | 8 (35) | 26 (27) | 7 (27) | 4 (40) | 46 (29) |
No time | 0 | 0 | 4 (17) | 19 (20) | 7 (27) | 1 (10) | 31 (20) |
Had physical contact with ≥1 other members | |||||||
Day before illness onset | 2 (100) | 2 (100) | 15 (75) | 69 (73) | 19 (73) | 5 (56) | 112 (73) |
Day before enrollment | 2 (100) | 1 (50) | 6 (32) | 37 (46) | 3 (14) | 2 (22) | 51 (38) |
Slept in the same room with ≥1 other members | |||||||
Day before illness onset | 0 | 1 (50) | 3 (15) | 43 (45) | 12 (46) | 5 (50) | 64 (41) |
Day before enrollment | 0 | 0 | 1 (5) | 29 (36) | 5 (24) | 2 (22) | 37 (28) |
Frequency of masking when interacting with ≥1 other members | |||||||
Day before illness onset | |||||||
Never | 1 (100) | — | 7 (88) | 44 (98) | 15 (83) | 9 (100) | 76 (94) |
Sometimes | 0 | — | 1 (13) | 1 (2) | 3 (17) | 0 | 5 (6) |
Always | 0 | — | 0 | 0 | 0 | 0 | 0 |
Day before enrollment | |||||||
Never | 1 (100) | — | 5 (63) | 33 (79) | 11 (65) | 7 (78) | 57 (74) |
Sometimes | 0 | — | 2 (25) | 3 (7) | 1 (6) | 0 | 6 (8) |
Always | 0 | — | 1 (13) | 6 (14) | 5 (29) | 2 (22) | 14 (18) |
. | Age, y . | ||||||
---|---|---|---|---|---|---|---|
0–4 . | 5–11 . | 12–17 . | 18–49 . | 50–64 . | ≥65 . | All Ages . | |
N (row %) | 2 (0.9) | 4 (1.8) | 25 (11.1) | 141 (62.4) | 37 (16.4) | 17 (7.5) | 226 (100) |
Female | 2 (100) | 0 (0) | 15 (60) | 81 (57) | 21 (57) | 9 (53) | 128 (56.6) |
Race/ethnicitya | |||||||
Non-Hispanic White | b | b | 23 (92) | 115 (82) | 35 (95) | 17 (100) | 194 (85.8) |
Non-Hispanic other race | b | b | 0 | 9 (6) | 1 (3) | 0 | 11 (4.9) |
Hispanic | b | b | 2 (8) | 16 (11) | 1 (3) | 0 | 20 (8.9) |
Smoker (age ≥18 y) | — | — | — | 7 (5) | 2 (5) | 0 | 9 (5) |
Any underlying medical conditionsc | 1 (50) | 0 | 9 (36) | 46 (33) | 13 (35) | 12 (71) | 81 (35.8) |
Occupational, school, or child care exposures | |||||||
Attended child care or school outside homed (age <18 y) | 1 (100) | 3 (100) | 14 (74) | — | — | — | 18 (78) |
Health care setting (age ≥18 y)d,e | — | — | — | 33 (24) | 6 (16) | 0 | 39 (20) |
Customer service (age ≥18 y)d,e | — | — | — | 31 (22) | 4 (11) | 2 (12) | 37 (19) |
Teacher (age ≥18 y)d | — | — | — | 4 (4) | 2 (7) | 1 (8) | 7 (5) |
Household characteristics | |||||||
Mean (SD) number of household membersf | 4.0 (0) | 4.3 (1.7) | 4.4 (0.9) | 3.4 (1.4) | 2.7 (0.8) | 2.1 (0.2) | 3.3 (1.4) |
Type of home | |||||||
Single family home | 2 (100) | 3 (75) | 25 (100) | 110 (78) | 33 (89) | 14 (82) | 187 (82.7) |
Duplex/townhome | 0 | 0 | 0 | 0 | 0 | 8 (3.5) | |
Condominium/apartment building | 0 | 1 (25) | 0 | 23 (16) | 4 (11) | 3 (18) | 31 (13.7) |
Mean (SD) number of bedroomsf | 4.0 (0) | 3.3 (1.0) | 3.9 (0.7) | 3.0 (1.0) | 3.4 (1.2) | 2.9 (0.7) | 3.2 (1.0) |
Interactions with other household membersd,g | |||||||
Maximum time spent in same room with ≥1 other member | |||||||
Day before illness onset | |||||||
>4 h | 2 (100) | 2 (100) | 10 (43) | 56 (59) | 10 (40) | 6 (60) | 86 (55) |
1–4 h | 0 | 0 | 7 (30) | 23 (24) | 11 (44) | 1 (10) | 42 (27) |
<1 h | 0 | 0 | 3 (13) | 16 (17) | 4 (16) | 3 (30) | 26 (17) |
No time | 0 | 0 | 3 (13) | 0 | 0 | 0 | 3 (2) |
Day before enrollment | |||||||
>4 h | 2 (100) | 1 (50) | 4 (17) | 35 (37) | 6 (23) | 2 (20) | 50 (32) |
1–4 h | 0 | 0 | 7 (30) | 15 (16) | 6 (23) | 3 (30) | 31 (20) |
<1 h | 0 | 1 (50) | 8 (35) | 26 (27) | 7 (27) | 4 (40) | 46 (29) |
No time | 0 | 0 | 4 (17) | 19 (20) | 7 (27) | 1 (10) | 31 (20) |
Had physical contact with ≥1 other members | |||||||
Day before illness onset | 2 (100) | 2 (100) | 15 (75) | 69 (73) | 19 (73) | 5 (56) | 112 (73) |
Day before enrollment | 2 (100) | 1 (50) | 6 (32) | 37 (46) | 3 (14) | 2 (22) | 51 (38) |
Slept in the same room with ≥1 other members | |||||||
Day before illness onset | 0 | 1 (50) | 3 (15) | 43 (45) | 12 (46) | 5 (50) | 64 (41) |
Day before enrollment | 0 | 0 | 1 (5) | 29 (36) | 5 (24) | 2 (22) | 37 (28) |
Frequency of masking when interacting with ≥1 other members | |||||||
Day before illness onset | |||||||
Never | 1 (100) | — | 7 (88) | 44 (98) | 15 (83) | 9 (100) | 76 (94) |
Sometimes | 0 | — | 1 (13) | 1 (2) | 3 (17) | 0 | 5 (6) |
Always | 0 | — | 0 | 0 | 0 | 0 | 0 |
Day before enrollment | |||||||
Never | 1 (100) | — | 5 (63) | 33 (79) | 11 (65) | 7 (78) | 57 (74) |
Sometimes | 0 | — | 2 (25) | 3 (7) | 1 (6) | 0 | 6 (8) |
Always | 0 | — | 1 (13) | 6 (14) | 5 (29) | 2 (22) | 14 (18) |
Data are no. (% of column total with response) unless otherwise noted, numbers reflect rounding. —, data not reported by participants.
Unknown for 1 participant.
Data suppressed to protect privacy.
Underlying medical conditions included: asthma, cancer, cardiovascular or heart disease, diabetes, extreme obesity, high blood pressure or hypertension, immunocompromising condition, kidney disease, liver disease, other chronic lung disease, pregnancy, and prematurity.
Missing responses.
Health care setting were those who reported working in a health care setting and having regular face-to-face contact with sick people. Customer service were those who reported working in customer service where they have regular face-to-face contact with people.
Data presented in this row are mean and SD.
Restricted to primary cases who were also the index case in the household.
Children age 0 to 4 years and adults age ≥65 years represented a minority (5% and 6%, respectively) of household contacts (Table 2). Other characteristics were similar to those reported among primary cases.
. | Age, y . | ||||||
---|---|---|---|---|---|---|---|
0–4 . | 5–11 . | 12–17 . | 18–49 . | 50–64 . | ≥65 . | All Ages . | |
N (row %) | 21 (5.2) | 53 (13.1) | 67 (16.6) | 184 (45.5) | 53 (13.1) | 26 (6.4) | 404 (100) |
Female | 9 (43) | 25 (47) | 28 (42) | 98 (53) | 28 (53) | 13 (50) | 201 (49.8) |
Race/Ethnicity | |||||||
Non-Hispanic White | 19 (90) | 40 (75) | 55 (82) | 148 (80) | 48 (91) | 24 (92) | 334 (82.7) |
Non-Hispanic other race | 1 (5) | 2 (4) | 4 (6) | 19 (10) | 1 (2) | 2 (8) | 29 (7.2) |
Hispanic | 1 (5) | 11 (21) | 8 (12) | 17 (9) | 4 (8) | 0 | 41 (10.2) |
Smoker (age ≥18 y) | — | — | — | 12 (7) | 3 (6) | 0 | 15 (6) |
Any underlying medical conditionsa | 0 | 6 (11) | 15 (22) | 53 (29) | 17 (32) | 16 (62) | 107 (26.5) |
Occupational, school, or child care exposures | |||||||
Attended child care or school outside homeb (age <18 y) | 5 (56) | 21 (66) | 29 (71) | — | — | — | 55 (67) |
Health care setting (age ≥18 y)b,c | — | — | — | 17 (9) | 5 (9) | 1 (4) | 23 (8.8) |
Customer service (age ≥18 y)b,c | — | — | — | 36 (20) | 10 (19) | 2 (8) | 48 (18.5) |
Teacher (age ≥18 y)b | — | — | — | 8 (6) | 2 (5) | 1 (5) | 11 (5.5) |
. | Age, y . | ||||||
---|---|---|---|---|---|---|---|
0–4 . | 5–11 . | 12–17 . | 18–49 . | 50–64 . | ≥65 . | All Ages . | |
N (row %) | 21 (5.2) | 53 (13.1) | 67 (16.6) | 184 (45.5) | 53 (13.1) | 26 (6.4) | 404 (100) |
Female | 9 (43) | 25 (47) | 28 (42) | 98 (53) | 28 (53) | 13 (50) | 201 (49.8) |
Race/Ethnicity | |||||||
Non-Hispanic White | 19 (90) | 40 (75) | 55 (82) | 148 (80) | 48 (91) | 24 (92) | 334 (82.7) |
Non-Hispanic other race | 1 (5) | 2 (4) | 4 (6) | 19 (10) | 1 (2) | 2 (8) | 29 (7.2) |
Hispanic | 1 (5) | 11 (21) | 8 (12) | 17 (9) | 4 (8) | 0 | 41 (10.2) |
Smoker (age ≥18 y) | — | — | — | 12 (7) | 3 (6) | 0 | 15 (6) |
Any underlying medical conditionsa | 0 | 6 (11) | 15 (22) | 53 (29) | 17 (32) | 16 (62) | 107 (26.5) |
Occupational, school, or child care exposures | |||||||
Attended child care or school outside homeb (age <18 y) | 5 (56) | 21 (66) | 29 (71) | — | — | — | 55 (67) |
Health care setting (age ≥18 y)b,c | — | — | — | 17 (9) | 5 (9) | 1 (4) | 23 (8.8) |
Customer service (age ≥18 y)b,c | — | — | — | 36 (20) | 10 (19) | 2 (8) | 48 (18.5) |
Teacher (age ≥18 y)b | — | — | — | 8 (6) | 2 (5) | 1 (5) | 11 (5.5) |
Data are no. (% of column total with response) unless otherwise noted, numbers reflect rounding. —, data not reported by participants.
Underlying medical conditions included: asthma, cancer, cardiovascular or heart disease, diabetes, extreme obesity, high blood pressure or hypertension, immunocompromising condition, kidney disease, liver disease, other chronic lung disease, pregnancy, and prematurity.
Missing responses.
Health care setting were those who reported working in a health care setting and having regular face-to-face contact with sick people, customer service were those who reported working in customer service where they have regular face-to-face contact with people.
Transmission From Primary Cases and Infection Among Contacts
The 226 primary cases were followed by 198 (49%) SARS-CoV-2 infections among 404 household contacts. At least 1 contact was infected in 58% (130 of 226) of households. Estimated SIR ranged from 26% among contacts of primary cases age 12 to 17 years to 76% among contacts of primary cases age ≥65 years (Fig 2). Compared to when the primary case was age 18 to 49 years, SIR in household contacts was significantly lower when the primary case was age 12 to 17 years (RR, 0.42; 95% CI, 0.19–0.91), and not significantly different for all other primary case age groups. There were no significant differences in estimated SIR by age of the contacts (Fig 2). SIR ranged from 36% among contacts age ≥65 years to 53% among contacts age 5 to 11 years.
Overall, estimated SIR was higher when primary case-contact pairs were in the same versus different age groups (55% versus 45%; RR, 1.47; 95% CI, 0.98–2.22; Fig 2). SIR was highest among primary case-contacts pairs age ≥65 years (76%) and 5 to 11 years (69%). Within each primary case age group, SIR was generally lowest among contacts age ≥65 years (Fig 3). However, CIs were wide for all primary case-contact age group combinations, particularly those age <12 years.
Symptoms Among Contacts With SARS-CoV-2 Infection
Among 186 contacts with SARS-CoV-2 infection with symptom onset or first positive rRT-PCR result after enrollment, 96% completed symptom surveys on ≥13 days of follow-up (Table 3). Most (81%) reported ≥1 symptoms and 19% reported no symptoms (asymptomatic) (Fig 4). There were no differences in asymptomatic infections by age group (ranging from 12% among those age 50-64 years to 27% among those age 5 to 11 years). Among the 150 symptomatic infections, 46% were rRT-PCR positive (from nasal or saliva sample) before symptom onset. Median days from first positive viral detection to symptom onset was 2 (interquartile range [IQR] 1–3) and did not differ by age group (P = .90). Median number of days with a positive rRT-PCR result (from nasal samples) during follow-up was 8 (IQR 4–11); however, 34% continued to be positive on the last day of follow-up.
. | Age . | ||||||
---|---|---|---|---|---|---|---|
0–4 . | 5–11 . | 12–17 . | 18–49 . | 50–64 . | ≥65 . | All Ages . | |
N (row %) | 8 (4) | 26 (14) | 30 (16) | 81 (44) | 25 (13) | 16 (9) | 186 (100) |
Female, n (column %) | 3 (38) | 11 (42) | 15 (50) | 36 (44) | 11 (44) | 10 (63) | 86 (46) |
Race/ethnicity, n (column %) | |||||||
Non-Hispanic white | 8 (100) | 20 (77) | 24 (80) | 63 (78) | 23 (92) | 16 (100) | 154 (83) |
Non-Hispanic other race | 0 | 0 | 2 (7) | 8 (10) | 0 | 0 | 10 (5) |
Hispanic | 0 | 6 (23) | 4 (13) | 10 (12) | 2 (8) | 0 | 22 (12) |
Smoker (age ≥18 y), n (column %) | — | — | — | 2 (2) | 1 (4) | 0 | 3 (2) |
Any underlying medical conditions,an (column %) | 0 | 2 (8) | 1 (3) | 20 (25) | 9 (36) | 10 (63) | 42 (23) |
Occupational, school, or child care exposures, n (column %) | |||||||
Attended child care or school outside homeb (age <18 y) | 2 (50) | 15 (65) | 11 (55) | — | — | — | 28 (60) |
Health care setting (age ≥18 y)b,c | — | — | — | 6 (8) | 3 (12) | 1 (6) | 10 (8) |
Customer service (age ≥18 y)b,c | — | — | — | 14 (18) | 4 (16) | 2 (13) | 20 (17) |
Teacher (age ≥18 y)b | — | — | — | 4 (6) | 1 (5) | 1 (8) | 6 (6) |
Completed ≥13 (of 14) symptom surveys, n (column %) | 8 (100) | 26 (100) | 29 (97) | 80 (99) | 24 (96) | 12 (75) | 179 (96) |
Took medication for fever or pain at least 1 d, n (column %) | 6 (75) | 9 (35) | 11 (37) | 41 (51) | 18 (72) | 9 (56) | 94 (51) |
Sought medical care during follow-up, n (column %)b | 0 | 0 | 2 (7) | 5 (6) | 0 | 2 (13) | 9 (5) |
Timing of positive result relative to symptom onset | |||||||
Positive rRT-PCR result before symptom onset, n (column %)d,f | 3 (43) | 11 (58) | 10 (42) | 30 (46) | 11 (50) | 4 (31) | 69 (46) |
Median (IQR) days from first positive to symptom onsetd,e,f | 2 (1– 3) | 3 (1– 3) | 2 (1– 3) | 1.5 (1–3) | 2 (1–4) | 2 (1–5) | 2 (1–3) |
Median (IQR) days positive during follow-up periodg | 9 (6.5–11) | 8 (3.5–10.5) | 8 (6–11) | 8 (4–11) | 12 (7–13) | 8 (1–12) | 8 (4–11) |
Median (IQR) days from primary case onset | |||||||
To first positive | 7 (6.5–9) | 6 (5– 9) | 6 (5– 8) | 6 (4–8) | 5 (5–7) | 5 (4–5.5) | 6 (4–8) |
To first symptomd | 8 (6– 8) | 6 (5– 10) | 6 (4.5–9) | 6 (4–7) | 6 (5– 7) | 5 (4– 6) | 6 (5– 8) |
. | Age . | ||||||
---|---|---|---|---|---|---|---|
0–4 . | 5–11 . | 12–17 . | 18–49 . | 50–64 . | ≥65 . | All Ages . | |
N (row %) | 8 (4) | 26 (14) | 30 (16) | 81 (44) | 25 (13) | 16 (9) | 186 (100) |
Female, n (column %) | 3 (38) | 11 (42) | 15 (50) | 36 (44) | 11 (44) | 10 (63) | 86 (46) |
Race/ethnicity, n (column %) | |||||||
Non-Hispanic white | 8 (100) | 20 (77) | 24 (80) | 63 (78) | 23 (92) | 16 (100) | 154 (83) |
Non-Hispanic other race | 0 | 0 | 2 (7) | 8 (10) | 0 | 0 | 10 (5) |
Hispanic | 0 | 6 (23) | 4 (13) | 10 (12) | 2 (8) | 0 | 22 (12) |
Smoker (age ≥18 y), n (column %) | — | — | — | 2 (2) | 1 (4) | 0 | 3 (2) |
Any underlying medical conditions,an (column %) | 0 | 2 (8) | 1 (3) | 20 (25) | 9 (36) | 10 (63) | 42 (23) |
Occupational, school, or child care exposures, n (column %) | |||||||
Attended child care or school outside homeb (age <18 y) | 2 (50) | 15 (65) | 11 (55) | — | — | — | 28 (60) |
Health care setting (age ≥18 y)b,c | — | — | — | 6 (8) | 3 (12) | 1 (6) | 10 (8) |
Customer service (age ≥18 y)b,c | — | — | — | 14 (18) | 4 (16) | 2 (13) | 20 (17) |
Teacher (age ≥18 y)b | — | — | — | 4 (6) | 1 (5) | 1 (8) | 6 (6) |
Completed ≥13 (of 14) symptom surveys, n (column %) | 8 (100) | 26 (100) | 29 (97) | 80 (99) | 24 (96) | 12 (75) | 179 (96) |
Took medication for fever or pain at least 1 d, n (column %) | 6 (75) | 9 (35) | 11 (37) | 41 (51) | 18 (72) | 9 (56) | 94 (51) |
Sought medical care during follow-up, n (column %)b | 0 | 0 | 2 (7) | 5 (6) | 0 | 2 (13) | 9 (5) |
Timing of positive result relative to symptom onset | |||||||
Positive rRT-PCR result before symptom onset, n (column %)d,f | 3 (43) | 11 (58) | 10 (42) | 30 (46) | 11 (50) | 4 (31) | 69 (46) |
Median (IQR) days from first positive to symptom onsetd,e,f | 2 (1– 3) | 3 (1– 3) | 2 (1– 3) | 1.5 (1–3) | 2 (1–4) | 2 (1–5) | 2 (1–3) |
Median (IQR) days positive during follow-up periodg | 9 (6.5–11) | 8 (3.5–10.5) | 8 (6–11) | 8 (4–11) | 12 (7–13) | 8 (1–12) | 8 (4–11) |
Median (IQR) days from primary case onset | |||||||
To first positive | 7 (6.5–9) | 6 (5– 9) | 6 (5– 8) | 6 (4–8) | 5 (5–7) | 5 (4–5.5) | 6 (4–8) |
To first symptomd | 8 (6– 8) | 6 (5– 10) | 6 (4.5–9) | 6 (4–7) | 6 (5– 7) | 5 (4– 6) | 6 (5– 8) |
IQR, interquartile range; rRT-PCR, real-time reverse transcription polymerase chain reaction. —, data not reported by participants.
Underlying medical conditions included: asthma, cancer, cardiovascular or heart disease, diabetes, extreme obesity, high blood pressure or hypertension, immunocompromising condition, kidney disease, liver disease, other chronic lung disease, pregnancy, and prematurity.
Missing responses.
Health care setting were those who reported working in a health caresetting and having regular face-to-face contact with sick people, customer service were those who reported working in customer service where they have regular face-to-face contact with people.
Among contacts who reported any symptoms. Symptoms assessed included: constitutional symptoms (chills, fatigue or feeling run down, fever or feverishness, muscle or body aches), upper respiratory symptoms (nasal congestion, runny nose, sore throat), lower respiratory symptoms (chest tightness or pain, cough, trouble breathing or shortness of breath, wheezing), neurologic symptoms (headache, loss of taste or smell), and gastrointestinal symptoms (abdominal pain, diarrhea, vomiting). Gastrointestinal symptoms were assessed in Wisconsin only.
Among contacts with a positive rRT-PCR result before symptom onset.
Impacted by left-censoring, participants may have been positive before their symptom onset and positive before the start of follow-up.
Restricted to nasal samples only; impacted by censoring (left and right).
Among infected children, the most commonly reported symptoms were upper respiratory symptoms (88% age 0 to 4 years, 62% age 5 to 11 years, and 77% age 12 to 17 years; Fig 4). Lower respiratory symptoms were reported by 68% (28 of 41) of adults ≥50 years of age (76% age 50 to 64 years, 56% age ≥65 years). Gastrointestinal symptoms were less common than other symptoms (40%) and not typically reported on the day of illness onset, but was commonly reported in children age 5 to 11 years (50%) and adults age 18 to 49 years (60%). Frequency and duration of individual symptoms are shown in Supplemental Fig 1.
Median duration of any symptom was 8 days (IQR 3 to 11) and did not differ by age group; 34% reported symptoms on the last day of follow-up, and thus their symptom duration was foreshortened. However, median duration of constitutional symptoms was significantly different and increased with age, from 2 days in younger children (age <12 years) to 9 days in adults ≥65 years of age (P = .03). Median duration of neurologic symptoms was ≤4 days for all age groups except adults age 50 to 64 years, where a median of 10 days was observed (P = .002).
Discussion
This prospective study examined the association of age with household transmission of SARS-CoV-2, addressing both the age of the primary case (transmission risk) and ages of household contacts (susceptibility). With frequent and systematic testing among household contacts, we found much higher infection risks among household contacts than in previous studies, similar rates of infection among child and adult household contacts, and some nuanced differences in transmission from specific age groups. However, transmission to household contacts was observed from primary cases in all age groups.
There is considerable heterogeneity in the literature on transmission from and susceptibility of children to SARS-CoV-2. Our finding of similar infection risk across age groups is consistent with larger investigations of reported cases from Spain,20 China,6 and Brunei,21 and other household transmission studies conducted in the United States.22 However, several studies have reported that children had reduced susceptibility to SARS-CoV-2 infection23–26 and one found higher risk in children compared to young adults.27 Difference in methods for ascertaining and detecting SARS-CoV-2 infection among household contacts may account for some of the differences between studies. Infrequent or delayed testing, or limited testing of asymptomatic or mildly-symptomatic individuals could contribute to underestimation of SARS-CoV-2 infections and may differentially impact infection risk measurements among children, thus biasing the association between age and susceptibility and transmission.
Additionally, some previous studies did not disaggregate age into finer age groups, analyzing all children <18 years of age together,5,9,22,26,28–31 which may obscure biological or behavioral characteristics that vary by age. We categorized our study participants into 6 age groups, roughly aligned with US groupings in school (preschool, primary school, and secondary school) and adulthood periods (young, middle-aged, and older adults). In doing so, we found that transmission risk was highest from primary cases ≥65 years of age and lowest from primary cases age 12 to 17 years. Studies from Ontario and Denmark also found transmission risk in children was highest among the youngest and lowest in adolescents.32,33 Age-specific differences in behavior likely contributed to lower transmission from adolescents relative to younger, less autonomous children.
Other studies have also identified somewhat increased transmission risk among contacts of similar age, and this may have implications beyond the household.33,34 Transmission patterns may be influenced by assortative mixing, where people of similar ages interact more with each other than with people of different ages.35,36 These findings may inform planning for school-based countermeasures to reduce transmission risk within and between classrooms. However, further investigation is needed to better understand how behaviors and interactions differ by age, during periods of illness, and whether those differences are associated with risk of transmission in both the household and school settings.
In this study and others,1 young children <12 years of age were rarely identified as the primary case of SARS-CoV-2 infection in households. However, when a young child was the primary case, we saw that they transmitted infection to >37% of their household contacts, including adults, and their probability of transmission in the household setting was not significantly different from transmission from young adults. Similar rates of transmissibility between children and adults were also found in studies conducted in China and South Korea,23,37 although small sample size in young children also limited those studies.
The frequency and duration of symptoms were similar across age groups among those who were infected in our study. Clinical data from nonhospitalized infected children are limited, with few studies directly comparing symptoms of children and adults.5,22,38,39 There were no significant differences in the frequency of asymptomatic infections between children and adults. The overall percentage of infected participants reporting no symptoms is generally consistent with the 20% estimated in a recent review and meta-analysis.40 Additional studies to better understand the frequency of asymptomatic infections by age and the role of asymptomatic infection and age in onward transmission in household and community settings are needed to inform public health recommendations.
This study has several limitations. First, delayed identification of index cases prevented complete capture of transmission events. Participants may have become infected but remained asymptomatic, between the time the index case was tested and study enrollment. Thus, duration of positivity and symptoms captured during the enrollment period may be subject to both right and left censoring and may be an underestimate. Second, we assumed secondary infections among contacts resulted from household exposure rather than community transmission. Ongoing exposure from the community may lead to an overestimation of transmission in household settings, especially among age groups more likely to be exposed outside the household. Although patients were instructed to isolate at home or quarantine while waiting for results,41 compliance may have declined as the pandemic progressed. Third, we did not account for reported interactions between the primary case and contacts before and during illness in the primary case. While physical contact between the primary case and household contacts was common, mask use by the primary case was not common before or after illness onset. Differences in age-related interactions may explain or help clarify associations between age and transmission events. Fourth, the study population was racially homogenous with low household density; results may not be generalizable to more diverse populations or more crowded household settings. Fifth, small sample size limited the precision of estimates of SIR and our statistical power to detect true differences in transmission risk and symptom profiles by age. Specifically, <14% of primary cases were children. Finally, this study was conducted before widespread circulation of the Delta variant, and transmission frequency may differ for Delta or other new variants. Despite these limitations, the case-ascertained household study described provides much needed evidence regarding susceptibility to SARS-CoV-2 infection, because exposure within the household is well-defined and prolonged. Additionally, daily follow-ups allowed us to assess symptoms throughout the course of illness.
We observed that both children and adults of all ages can transmit and are susceptible to SARS-CoV-2 infection. There were no significant differences in susceptibility to SARS-CoV-2 by age group, from preschool-aged children to older adults. Further research is needed to understand age-related interactions and behaviors in households as it relates to the probability of transmission by age.
Acknowledgments
We thank the following for their contributions to the study: Lynn Ivacic, Hannah Berger, Vicki Moon, Keegan Brighton, Gina Burbey, Deanna Cole, Leila Deering, Eric DeJarlais, Heather Dirkx, Sherri Guzinski, Joshua Hebert, Linda Heeren, Erin Higdon, Jacob Johnston, Chris Kadolph, Taylor Kent, Burney Kieke, Tamara Kronenwetter Koepel, Sarah Kohn, Diane Kohnhorst, Erik Kronholm, Stacey Kyle, Jim Linneman, Carrie Marcis, Karen McGreevey, Sudha Medabalimi, Nidhi Mehta, Nan Pan, Cory Pike, Rebecca Pilsner, DeeAnn Polacek, Martha Presson, Carla Rottscheit, Jacklyn Salzwedel, Kristin Seyfert, Tapan Sharma, Alyssa Spoerl, Sandy Strey, Krishna Chaitanya Upadhyay, Gail Weinand, and Benjamin Zimmerman at Marshfield Clinic Research Institute; Judy King, Dayna Wyatt, Robert Lyons, Carleigh Frazier, Emily Jookar, Karen Malone, Olivia Doak, Sarah Davis, Jorge Celedonio, Marcia Blair, Rendie McHenry, Claudia Guevara, Jennifer Luther, Laura Short, and Ahra Kim at Vanderbilt University Medical Center.
Dr McLean conceptualized and designed the study, designed the data collection instruments, coordinated and supervised the data collection, carried out the analyses, drafted the initial manuscript, and critically reviewed the manuscript for important intellectual content; Drs Grijalva, Belongia, Talbot, and Rolfes conceptualized and designed the study, designed the data collection instruments, coordinated and supervised the data collection, and critically reviewed the manuscript for important intellectual content; Ms Hanson designed the data collection instruments, coordinated and supervised data collection, and critically reviewed the manuscript for important intellectual content; Drs Zhu, Meece, Halasa, and Chappell coordinated and supervised the data collection and critically reviewed the manuscript for important intellectual content; Ms Deyoe coordinated data collection and critically reviewed the manuscript for important intellectual content; Dr Mellis critically reviewed the manuscript for important intellectual content; Dr Reed conceptualized and designed the study and critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the US Centers for Disease Control and Prevention.
FUNDING: This study was supported by the Centers for Disease Control and Prevention, (cooperative agreements IP001078 and IP001083). Dr Grijalva was supported in part by the National Institute for Allergy and Infectious Diseases (grant K24 AI148459). The work used REDCap, which is supported by CTSA award No. UL1 TR002243 from the National Center for Advancing Translational Sciences. Investigators at the Centers for Disease Control and Prevention contributed to the design and conduct of the study.
- CDC
Centers for Disease Control and Prevention
- CI
confidence interval
- COVID-19
coronavirus disease 2019
- IQR
interquartile range
- MCHS
Marshfield Clinic Health System
- RR
risk ratio
- rT-PCR
real-time reverse transcription polymerase chain reaction
- SARS-CoV-2
severe acute respiratory syndrome coronavirus 2
- SIR
secondary infection risk
- VUMC
Vanderbilt University Medical Center
References
Competing Interests
CONFLICT OF INTEREST DISCLOSURES: Dr Halasa reports grants from Sanofi and Quidel. Dr Grijalva reports grants from Campbell Alliance/Syneos, the National Institutes of Health, the Food and Drug Administration, the Agency for Health Care Research and Quality and Sanofi-Pasteur, and consultation fees from Pfizer, Merck, and Sanofi-Pasteur. Other authors have no conflicts of interest relevant to this article to disclose.
Comments