For a week or so, there wasn’t a lot of interesting new research being dropped at preprint servers. Now all of a sudden it exploded, so it will take me a couple of posts to begin to catch up.
If you are interested in more information about the various kinds of testing for coronavirus, this is a short but very helpful article. (JAMA Article) Both infection and antibody testing are described.
This paper looked at an important issue–comparison of viral load across patients and with other viruses. (Medrxiv Paper) Over 10,000 specimens were studied, including 4,000 coronavirus ones. Viral loads for coronavirus were similar to other viruses. There was little difference across age or gender. Children rarely tested positive, but had similar viral loads. Virus load tended to be higher earlier in the epidemic. Viral loads tended to be higher earlier in the disease course in an individual. The viral load data could help determine not only how a patient is responding to the disease, but level of infectiousness.
Just because someone tests positive on an infection test, doesn’t mean they are infectious, that is, that they are capable of transmitting live virus in sufficient quantities to infect another person. This study compared positive infection test results with cell culture findings. (Medrxiv Paper) Samples were collected, some more than once from the same patient, from 178 outpatients, 12 inpatients and 5 patients in the ICU. By culture, only 15% of the outpatient samples had viable virus, 45% of the inpatient ones did, and 82% of the ICU ones. The average time from symptom duration to a positive culture test was 4.5 days. Only one patient was still culture positive longer than ten days after symptoms appeared.
This study tracked real-time health care worker exposure to infected patients to determine if much airborne transmission was occurring. The authors found little evidence for such transmission, less than with other respiratory viruses. (Medrxiv Paper) The analysis indicating that strong ventilation appeared to lessen the risk of airborne transmission.
I have struggled to understand a potential seasonality pattern of this strain of coronavirus. If there is one, it could be quite regional. This paper looks further at the seasonality of the common coronavirus strains. (Medrxiv Paper) The authors examined over 1250 respiratory tract samples, about 288 positive for coronavirus. Seasonality likely contributes to variation in an epidemic through weather, largely, and it may affect the virus, people’s behavior or certain characteristics of a person, like vitamin D levels. In addition to just time of year, the researchers looked at various meteorological measures. It should be noted that the samples came from Germany so there is not a lot of geographical variation. As other studies have indicated, the seasonal coronavirus infections peaked in December through March and were very low from July through September. Factors that possibly were associated with more infections were lower temperature, lower relative humidity, high cloud cover, and high precipitation. People who were immunocompromised in some manner showed less of a seasonal pattern. The authors hypothesized some potential reasons for the apparent associations, such as precipitation or low temperatures keeping people inside, more cloud cover may mean less ultraviolet radiation that kills viruses, and drier weather dries up nasal passages, hindering innate immune system work.
This is yet another paper attempting to describe the immune system responses to coronavirus. (Science Article) The researchers, from Penn University, did characterizations of the immune response in 149 hospitalized patients, 46 non-hospitalized ones and 70 healthy donors. Detailed information was available on a large majority of each of these groups. The hospitalized patients were quite a bit older and had high rates of comorbidities. Most of these hospitalized patients had robust T cell responses, but not all did. In general the mild cases and healthy people had similar immune system characterization. Some patients showed an excessive immune response that worsened the clinical presentation of their disease. About 75% of the hospitalized patients had immune responses that clustered in two main buckets, while the remainder had a response that looked like that of the milder patients. Overall, there appeared to be great variation in immune response, but those variations were related to the clinical course of the disease. Among other things, the work could lead to the ability to predict clinical manifestations based on the type of immune response.
People are looking for any kind of correlates of infection and serious disease. This research looked at blood types. (Medrxiv Paper) It compared the proportion of a population with certain blood types with infection and death rates. They said that more type A was associated with higher rates of cases and deaths. More type B was negatively associated. No causal explanation was attempted and I am dubious.
Finally, another seroprevalence study, this one from a very poor area in Buenos Aires, Argentina. (Medrxiv Paper) A very sensitive antibody test was used. Among those tested in a random sample of 577 households, with 873 people tested, over 53% were positive. Substantial clustering was noted, with twice the likelihood of having been infected if you lived in the household of someone else who was infected. Only 15% of those who were positive had experienced any symptoms. The prevalence was 9 times greater than that found by infection testing in the area.