In the first year of the SARS-CoV-2 pandemic, data were assembled that defined how the virus spread among us, who, at the time, had never been infected. But the world’s population is very different now – billions of infections, many waves triggered by evolved variants, vaccination, and that has impacted incubation periods, serial interval, generation times, and, according to this modelling, increased the proportion of cases (=infections) that are asymptomatic.
A study published last year applied a modelling system. It started by acknowledging that these estimates are challenged by many variables. These and other variables include:
- Testing policies
- Healthcare capability and accessibility
- Public awareness and engagement
- Ongoing virus evolution, especially to escape immunity
- Ongoing epidemics
- Changing vaccine formulations
The study also reviewed the literature and noted that previous investigations have reported asymptomatic cases at between 1.8% up to 45% of SARS-CoV-2 infections, with meta-analyses bringing the higher end down to 15-25%.
Modelling found two important trends over time
Sure, this is modelling, so of course, there are limitations, and every model is as good as what’s been considered and the data it was developed from. Please check the paper to see the limitations the authors already identified. You can add your own in the comments as well.
Below are the graphs and a summary table highlighting the modelled transmission rates and reproduction numbers (Rt).
The authors accounted for the variables in action during three different 6-month time periods (each divided into 6-month segments). In Period I, the early days, there was no vaccine (Sep 2020 – Feb 2021). Period II saw the vaccine rollout, coinciding with the Delta variant era (Jul 2021 – Dec 2021). Lastly, Period III contained many vaccinations and boosters coinciding with a predominance of Omicron variants (Jan 2022 – Jun 2022).
The plots show that at the transmission peak, the Rt (shown as lines) was predicted to be 1.8178 in Period I. The lowest Rt of 0.4909 occurred at the end of Period III. Average RTs didn’t differ much (Table 3), but what happened within the 6-month blocks of Period I-III was much more stark.
Pre-vaccine, transmission rates bounced around as various non-pharmaceutical measures were rolled out and withdrawn. Things stabilised in Period II as the vaccine was rolled out. In Period III, a more pronounced decline in Rt was evident over time as the combined effects of pre-existing immunity and new immunity from vaccination surges (including lots of boosting) and infections among the unboosted or unvaccinated slowed transmission.
Asymptomatic infections are likely to increase
The authors’ model also allowed them to propose that the probability of asymptomatic cases was higher as immunity grew. So, in Period III, there were likely more asymptomatic cases – esimated ot be 44.71% (Table 4) – than in the early days of the pandemic (31.55%).
Vaccination drives more asymptomatic cases
It probably follows that, in the vaccination-and-boosting timelines, with safer immunity than being unvaccinated and then infected, 49.70% of infections (Period III, Table 4) were estimated to be asymptomatic, compared to 41.70% in Period II.
Period I didn’t have any vaccine, so it wasn’t included.
Model built, then compared to real historical data
Okay, it was a model, so it’s fair to say that if you put garbage into a model’s development, you get garbage out. Also, swap “garbage” for “bias “.
In this instance, the authors built their model and then tested it by applying it during the conditions of the three periods to see if it successfully estimated what actually happened during those periods. And it did reproduce historical patterns with a good degree of similarity.
It’s still a model built with hindsight, so you’d hope it reflects the past, right? Is it truly predictive of the future? Ask an expert in these matters. If you are one, feel free to comment below.
Good news, bad news
Vaccines make a difference in many things. In this case, they prevent greater illness.
As a fully immunised (=successfully vaccinated and boosted) individual, this study is good news. If you get infected, you are more likely to come away with an unnoticed or perhaps milder case (not specifically studied here, though) of COVID-19. It does not deal with the issue of longCOVID.
But if you were not vaccinated or boosted, or if you were but had conditions that did not allow your system to produce a suitable immune response, then you were less likely to be part of the mild spectrum of disease. That means you are more likely to have an outcome that slides more towards the Moderate>Severe end of the disease spectrum.
Also, because we know that SARS-CoV-2 transmission starts before signs and symptoms of disease occur, this modelling implies (though it doesn’t show evidence for) that more occult (hidden, unseen) transmission will happen.
The best bang-for-buck way to reduce the spread of this and other harmful pathogens that transmit through the air is to clean that air. Exchange more of it for fresh outdoor air, filter it, or treat it. The next-best method is to wear tight-fitting, well-sealed N95 or P2 respirators (or even better-rated ones).
Sources
- Estimation of undetected asymptomatic infections of COVID-19: a mathematical modeling approach.
https://www.nature.com/articles/s41598-025-28374-y
Discover more from Virology Down Under
Subscribe to get the latest posts sent to your email.