Vitamin D, explaining R and the 2 metre rule

Episode Summary

Executive Summary: This episode examines how COVID-19 metrics are estimated and communicated, focusing on the reproduction number R, problems with government testing statistics, the science behind two-metre distancing, and tentative evidence about vitamin D. The core message is that key pandemic numbers are useful but uncertain, and official reporting can blur important distinctions.

Main Topics: Understanding the reproduction number (R) (Priority: 5/5): Tim Harford and epidemiologist Paul Birrell explain what R means, how it differs from a simple count of infections, and how it is estimated indirectly from deaths, hospital data, mobility, and contact patterns. Regional variation in R and model limits (Priority: 4/5): The discussion explores why London’s estimated R is lower than other regions and stresses that modelling relies on simplifying assumptions about behaviour, age, and contact patterns that may not fully match reality. Government COVID-19 testing statistics and target claims (Priority: 5/5): The program critiques the government’s repeated use of inflated testing totals that include posted tests, non-diagnostic tests, and multiple tests per person, arguing that the headline 100,000 target is not being met as claimed. Why the social distancing rule is two metres (Priority: 4/5): An infectious disease expert explains the physical basis of droplet spread, the difference between larger and smaller particles, and why risk decreases with distance, ventilation, and shorter exposure. Vitamin D and COVID-19 risk (Priority: 4/5): The episode reviews early observational studies suggesting a relationship between low vitamin D and worse COVID-19 outcomes, but cautions that the evidence is not strong enough to prove causation or justify dramatic claims. Need for better evidence and transparent metrics (Priority: 3/5): Across all segments, the episode emphasizes the importance of clear definitions, timely data, and randomized trials before making strong policy or medical claims.

Key Arguments: R is an average transmission measure: if R is above 1 the epidemic grows, and if below 1 it shrinks; it cannot be observed directly and must be inferred from lagged data. Estimated R below 1 means many infected people transmit to fewer than one person on average, not that each person infects a fractional number in practice. London’s lower R is attributed mainly to having experienced more of the first wave and therefore having fewer susceptible people left, not necessarily because behavior is more compliant. Regional R estimates are model-based and depend on simplifying assumptions about homogeneous behavior within age and region groups. Government testing claims are misleading because they count tests when they are posted, include non-diagnostic and antibody tests, and may count multiple tests on the same person. The government’s public testing numbers do not clearly show how many people were actually tested or how many postal tests were returned and processed. The two-metre rule is grounded in droplet physics, but actual infection risk depends on distance, duration, indoor/outdoor setting, air dilution, and masks. Vitamin D may plausibly help respiratory health, but the COVID-19 evidence discussed is observational and confounded, so it cannot establish that supplements prevent severe disease or death. Strong claims like being 19 or 20 times more likely to die from COVID-19 due to vitamin D deficiency require much stronger evidence than the studies cited. Randomized trials are needed before vitamin D can be recommended as a treatment for COVID-19 outcomes, though general supplementation advice may still be reasonable for some groups.

Data Points: Estimated R in England: 0.7 to 0.9 - Government believes R is below 1, implying epidemic decline Possible R in March: higher than 3 - Estimate just before lockdown Delay from infection to death: about 20 days - Used in back-calculation models from death data Testing target: 100,000 tests per day - Government benchmark repeatedly cited Posted tests included in one daily figure: more than 28,000 - Sunday/Monday testing totals included tests merely put in the post Additional non-diagnostic tests included: more than 17,500 - Tests from organizations like ONS and universities counted in completed tests May 10 tests carried out: almost 70,000 - Public figure for tests actually conducted that day May 10 people tested: just over 37,000 - Shows tests exceeded unique people by nearly 2:1 Clinical trials of vitamin D and respiratory infection: 41 trials - Systematic review cited by Adrian Martineau Participants in vitamin D respiratory infection review: more than 30,000 - Combined sample size across trials Overall protective effect of vitamin D: around 20% less likely - Meta-analysis on acute respiratory infections, not COVID-19 specifically Philippines observational study sample: 200 patients - Early study comparing vitamin D levels and COVID-19 severity Philippines study mild cases: 49 patients, 47 with normal vitamin D - Reported association between mild symptoms and normal vitamin D Philippines study critical cases: 48 patients, 2 with normal vitamin D - Reported association between critical symptoms and low vitamin D Second observational study sample: 780 cases - Compared deaths with hospitalized survivors Claimed increased odds of death with vitamin D deficiency: 19 times higher - Early preprint finding discussed with caution Public Health England supplement advice: 10 micrograms all year round - Recommended for people with little sun exposure and people with dark skin Distance guideline cited by CDC: 6 feet / 1.8 metres - International social distancing benchmark WHO-referenced closer benchmark: within 1 metre - Based on large droplet settling at shorter distances

Pivotal Quotes: "if R is more than one, the epidemic is growing, and if R is less than one, the epidemic is shrinking" — Tim Harford: Introduces the central epidemiological concept of the episode "No, no, they didn't. That figure, once again, included more than 28,000 tests which were simply put in the post that day." — Kate Lambell: Explaining why the government did not actually meet its testing target "These are early research studies, and I think the key thing to point out about them is that they're not clinical trials" — Adrian Martineau: Cautioning against overinterpreting vitamin D findings

Implications: Listeners should treat pandemic statistics carefully, distinguishing estimates from counts and headlines from definitions. Policy and health advice should rest on transparent data and stronger evidence, especially for testing metrics and proposed treatments like vitamin D.