Dr. Debra Soh on the Joe Rogan Experience #1520, fact-checked

What the evidence shows: Several small, early observational studies from 2020 did report that a large majority of hospitalized or critically ill COVID-19 patients had low vitamin D levels, and figures in that range (roughly 80-90% insufficient, single-digit percent sufficient) circulated in early-pandemic commentary, including from researchers such as Dr. Rhonda Patrick, whom Rogan credits as his source in this clip. The specific 80%/4% split cannot be tied to a single well-documented, widely replicated study. The broader body of evidence assembled since then is inconsistent: the NIH Office of Dietary Supplements' clinical summary notes that while some observational studies link lower vitamin D status to higher COVID-19 incidence and severity, others do not, and a meta-analysis of 31 observational studies found no significant association between low vitamin D and COVID-19 incidence, mortality, ICU admission, or need for ventilation. Randomized controlled trials designed to test whether correcting vitamin D levels actually improves COVID-19 outcomes have mostly failed to show benefit: per the same NIH summary, a 240-patient trial of high-dose vitamin D3 in hospitalized COVID-19 patients found no significant reduction in hospitalization length, mortality, ICU admission, or need for mechanical ventilation, and the French COVIT-TRIAL found an early mortality benefit at day 14 that had disappeared by day 28. Overall, the claim reflects a real pattern seen in early observational data but presents it as more settled and quantified than the evidence supports, and omits that later, larger observational analyses and randomized trials did not consistently confirm a protective effect of correcting vitamin D levels on COVID-19 severity.

What the evidence shows: Soh claims biological sex is binary because it is defined by gamete type, with no intermediate gametes, and that this applies to 99% of people. It is true that only two gamete types (ova and sperm) exist in humans and no third, intermediate gamete has been documented; a gamete-based definition of sex is a real position argued in the scientific literature, including a 2025 paper in Archives of Sexual Behavior titled "Why There Are Exactly Two Sexes," which has since drawn published rebuttals contesting its framing. However, mainstream biological and medical sources note that sex classification in practice draws on multiple traits, chromosomes, gonads, hormones, and anatomy, that do not always align neatly into two categories; conditions grouped under "differences of sex development" (DSDs) produce anatomical, chromosomal, or hormonal variation in an estimated 0.05% to 1.7% of the population depending on definition. Sex and gender are also distinct concepts in the biomedical and psychological literature, not interchangeable as the claim implies. The specific 99% figure was not sourced or defined by Soh during the discussion and does not correspond to a commonly cited statistic in the literature reviewed. Overall, gamete type is one valid, real biological framework for classifying sex, but presenting it as settling the sex/gender question for "99%" of people, without qualification, oversimplifies documented biological variation and elides the sex/gender distinction.

What the evidence shows: Soh cited a study finding that about 10% of millennials identify as a "third gender," including non-binary or genderqueer identities. The most authoritative and frequently cited survey on this exact question, Pew Research Center's 2022 nationally representative poll, found that 1.6% of all U.S. adults identify as transgender or nonbinary, rising to about 5% among adults under 30 (which includes only the youngest millennials, alongside Gen Z). Some smaller or non-representative online polls with broader or looser wording (e.g., counting anyone who selects a "genderqueer," "gender-fluid," or write-in category, or surveying self-selected online panels) have produced higher figures, which may be the source Soh is referencing, but these are not considered as methodologically robust as Pew's probability-based sample. Based on the best available nationally representative data, the specific 10% figure appears overstated for millennials as a generation.

What the evidence shows: No systematic review or meta-analysis pooling exactly 27 studies on reduced gender dysphoria and improved life satisfaction after transition could be located in PubMed, Europe PMC, or Crossref searches. The closest comparable and widely cited review, Baker et al. (2021, Journal of the Endocrine Society), examined 20 studies (reported across 22 publications) on gender-affirming hormone therapy and found it was associated with increased quality of life and decreased depression and anxiety, but the authors explicitly rated certainty in this conclusion as limited by high risk of bias, small sample sizes, and confounding with other interventions. Other systematic reviews in this area, including those commissioned for the UK's Cass Review, similarly describe the evidence on psychological and life-satisfaction outcomes after transition as low-to-moderate quality, drawn mostly from observational cohort studies rather than randomized trials. The general direction Soh describes, that some studies show improved outcomes after transition, is consistent with parts of the published literature, but the specific study count and the degree of certainty implied by citing it as a definitive meta-analysis are not substantiated by any identifiable single source.

What the evidence shows: Soh claimed there is no evidence that pornography addiction exists. The scientific and clinical picture is more mixed than that framing suggests. The World Health Organization's ICD-11, effective 2022, added "Compulsive Sexual Behaviour Disorder" as an official diagnosis, but deliberately classified it as an impulse-control disorder rather than a behavioral addiction, reflecting genuine disagreement among researchers over whether the addiction model best fits the condition. A 2023 systematic review of 20 studies applying DSM-5 substance-use-disorder criteria to problematic pornography and sexual behavior found that features such as craving, loss of control, and negative life consequences were highly prevalent among affected individuals, evidence that runs counter to a "no evidence" claim. At the same time, "pornography addiction" is not a formal DSM-5 diagnosis, and researchers remain divided on whether compulsive use reflects a true addictive process, a different clinical mechanism, or is better understood outside a disease framework altogether. Current status: the claim of "no evidence" overstates the case; the more accurate description is that the existence and mechanism of pornography addiction remain scientifically contested, with evidence on both sides rather than an absence of evidence.