Fig. 3: Prospective study of long COVID symptoms across the menstrual cycle.

A Boxplot representing the median number of daily symptoms across the three phases of the menstrual cycle. Box plots display the median (centre line), first quartile (Q1, lower box boundary) and third quartile (Q3, upper box boundary), with whiskers extending to the minimum and maximum values within 1.5×IQR of the box boundaries. Late Secretory/Menstrual (LS/M) phase: n = 32, minimum=3, Q1 = 9, median=13, Q3 = 18.25, maximum = 23; Proliferative (P) phase: n = 32, minimum=1, Q1 = 8, median=13.75, Q3 = 18.25, maximum = 23; Secretory (S) phase: n = 32, minimum=4, Q1 = 8.38, median=13.25, Q3 = 17, maximum = 22. The results of a two-sided Friedman test indicate no significant differences in the median number of symptoms between phases within individuals. B Density plot illustrating the probability distribution of symptom counts across the menstrual cycle phases. There was a large overlap in the distribution of the number of symptoms across the phases. C Distribution of symptom severity for each symptom across the phases of the menstrual cycle. D Predicted values and 95% confidence intervals for the association between menstrual cycle phase and the severity of long COVID-19 symptoms. Sample size: n = 54 participants (biological replicates) reporting across 29 symptoms, contributing 18,506 observations (technical replicates) over 930 tracking days. Menstrual phase distribution: S phase (n = 44 participants, 8143 observations); LS/M phase (n = 54 participants, 7751 observations); P phase (n = 33 participants, 2612 observations). Unit of study: individual participants with daily symptom tracking. Estimates derived from unadjusted univariable cumulative-link mixed models (one model per symptom) examining symptom severity across menstrual cycle phases. Cumulative odds ratios represent the odds of experiencing higher versus lower severity categories, with the secretory (S) phase as the reference category. Several symptoms showed significantly increased severity during specific menstrual phases: Late secretory/menstrual (LS/M) phase was associated with more severe dizziness (OR = 1.94, FDR-adjusted p = 0.005), fatigue (OR = 1.80, FDR-adjusted p = 0.0000), headache (OR = 2.11, FDR-adjusted p = 0.0000), muscle aches (OR = 1.48, FDR-adjusted p = 0.038), post-exertional malaise (OR = 1.45, FDR-adjusted p = 0.016), and tinnitus (OR = 1.42, FDR-adjusted p = 0.034) compared to the S phase. Proliferative (P) phase was associated with more severe breathing issues (OR = 3.14, FDR-adjusted p = 0.015), headache (OR = 2.81, FDR-adjusted p = 0.0000), post-exertional malaise (OR = 1.63, FDR-adjusted p = 0.024), with a trend toward more severe nausea (OR = 2.25, FDR-adjusted p = 0.055) and dizziness (OR = 1.83, FDR-adjusted p = 0.072) compared to the S phase. Red dots indicate statistically significant associations (Pr(>|z|) <0.05). Data and code are available on GitHub https://github.com/ataquette/Long-COVID-Mens.