New paper on children's viral loads (from symptomatic AND asymptomatic kids) & fraction asymptomatic, split out by 3 age groups (0-5, 6-13, 14-20).
Q thread
medrxiv.org/content/10.110…
Q thread
medrxiv.org/content/10.110…
Background
Still substantial uncertainty about how infectious children are, & still no estimates of age-specific fraction of infections that are asymptomatic.
Previous studies on viral load in kids came from symptomatic children (often very sick children) who might be anomalous.
Still substantial uncertainty about how infectious children are, & still no estimates of age-specific fraction of infections that are asymptomatic.
Previous studies on viral load in kids came from symptomatic children (often very sick children) who might be anomalous.
Measuring fraction w/ asymptomatic infections is key b/c it helps us interpret other data that are based on detecting symptomatic infection. Long discussion here:
New paper based on contacts of known cases (so avoids biases associated with detecting infections by symptoms). Sample size is moderate & critically includes viral loads asymptomatic children which some have argued might be lower (even though that's not the case in adults).
Fraction w/out symptoms varied among ages & was highest in 6-13; effect size moderate (25% vs 40%). Frac w/ respiratory symptoms or fever ~50% so likely to be missed 1/2 the time if focusing on just 1 of these (e.g. fever checks). 
Viral loads were similar across ages, similar b/w symptomatic & asymptomatic children, & showed normal decay over time after symptom onset. Study didn't report multiple regression to see if differential timing of sampling might have influenced loads among groups.
Finally, 1/3 of infected kids had infected siblings but no infected adult contact suggesting child-to-child transmission w/ infector being median 12 (IQR 8-16) yrs old.
While it's a relatively limited dataset (178 viral loads; symptoms in 293 kids), methods are ideal (subjects were close contacts of cases so not biased by symptom detection) & shouldn't be dismissed as anomalous data.
Major conclusions:
-Many (60-75%) children symptomatic but so mild could be missed, especially if based on 1 symptom; asymptomatic fraction highest in 6-13 & 50% higher than 0-5, 14-20.
-Higher fraction asymptomatic than meta-analysis (20%) of all ages medrxiv.org/content/10.110…
-Many (60-75%) children symptomatic but so mild could be missed, especially if based on 1 symptom; asymptomatic fraction highest in 6-13 & 50% higher than 0-5, 14-20.
-Higher fraction asymptomatic than meta-analysis (20%) of all ages medrxiv.org/content/10.110…
- Viral loads similar b/w ages w/in kids; supports earlier comparisons & makes it clear viral loads do not differ much b/w kids & adults
- Viral loads of asymptomatics same as symptomatics. We CANNOT dismiss past studies based on symptoms
- Viral loads of asymptomatics same as symptomatics. We CANNOT dismiss past studies based on symptoms
- Young siblings likely transmitted virus to each other, w/ 1/2 of infectors <=12 yrs old. Don't need a parent to be infected to have w/in household transmission.
Remaining unknowns:
- Study on children 0-20 yr so no comparisons can be made b/w kids & adults; but other papers provide bridge across ages & this paper addresses potential biases raised
- Viral loads are proxy for infectiousness; more unbiased contact tracing studies needed
- Study on children 0-20 yr so no comparisons can be made b/w kids & adults; but other papers provide bridge across ages & this paper addresses potential biases raised
- Viral loads are proxy for infectiousness; more unbiased contact tracing studies needed
-Schools not in session during study; will different contact types and rates during school produce similar or different results? Will masks be worn at school & will this be effective in reducing infection or infectious dose? Still uncertain if that is real phenomenon.
The lead author @hjillianh also has a nice detailed thread about the paper!
