Digital health app data reveals an effect of ovarian hormones on long COVID and myalgic encephalomyelitis symptoms

A retrospective analysis of digital health app data reveals that menstruation significantly increases the frequency of symptom crashes and severity in Long COVID and ME/CFS, while combined hormonal contraception is associated with reduced symptom burden, suggesting a modulatory role of ovarian hormones in these conditions.

The Gist

Imagine your body is like a high-performance car. For most people, this car runs smoothly every day. But for the millions of people living with Long COVID or ME/CFS (a condition often called chronic fatigue syndrome), that car has a very sensitive engine. It doesn't just break down; it has "crashes"—sudden, severe stalls where the car refuses to move, often triggered by doing too much.

This study is like a mechanic's logbook that looked at thousands of these "cars" to see if the seasons (the menstrual cycle) affect how well they run.

Here is the breakdown of what they found, using simple analogies:

1. The "Seasons" of the Body

Think of the menstrual cycle as having four distinct seasons:

• Menstruation (The Storm): The time of the period.
• Follicular (The Spring): The time after the period, leading up to ovulation.
• Luteal (The Summer/Autumn): The time after ovulation, leading up to the next period.
• Premenstrual (The Late Autumn): The few days right before the period starts.

The Finding: The study discovered that for people with Long COVID or ME/CFS, the "Storm" season (Menstruation) is the most dangerous time for their engines.

• The "Crash" Rate: Just like a car is more likely to stall in a hurricane, these patients were significantly more likely to have a severe symptom crash during their period than at any other time of the month.
• The "Safe Zone": The "Early Summer" (Early Luteal phase) was the safest time, where the engine ran the smoothest and crashes were least likely.

2. The "Stabilizer" (Hormonal Contraception)

The researchers also looked at people who use combined hormonal contraception (like the birth control pill, patch, or ring). You can think of these as "Hormonal Stabilizers."

• Without Stabilizers: The body's natural hormones go up and down like a rollercoaster. This wild ride seems to make the sensitive Long COVID/ME/CFS engine sputter and crash more often.
• With Stabilizers: The pill keeps the hormones on a flat, smooth road instead of a rollercoaster.
• The Result: People using these stabilizers reported fewer crashes and less severe symptoms overall. It's as if the pill acted like a shock absorber, smoothing out the bumps that usually trigger a breakdown.

3. The "Control Group" (The General Public)

To make sure this wasn't just in their heads, the researchers compared these patients to a group of healthy women who also tracked their cycles (but didn't have Long COVID).

• The Surprise: Even the healthy women felt a little more tired or had more headaches during their period. It's like how everyone feels a bit sluggish when it rains.
• The Difference: However, for the Long COVID/ME/CFS group, the "rain" turned into a flood. Their symptoms didn't just get a little worse; they got significantly worse, and the "crashes" (total shutdowns) were unique to them.

4. Why This Matters (The Takeaway)

This study is a game-changer because it moves from "guessing" to "knowing."

• For Patients: It's like getting a weather forecast for your own body. If you know your "Storm Season" is coming next week, you can plan ahead. You might decide to rest more, skip a big event, or save your energy for the "Safe Zone" (Early Summer) when you feel your best.
• For Doctors: It suggests that for some patients, using hormonal contraception might not just be about preventing pregnancy; it could be a treatment to help calm the symptoms of Long COVID and ME/CFS.
• The Big Picture: It confirms that sex hormones (estrogen and progesterone) are like the fuel mix for the immune system. When the fuel mix changes too wildly, the engine of these chronic illnesses misfires.

In a Nutshell

Think of the menstrual cycle as the tide. For most people, the tide rises and falls gently. For people with Long COVID and ME/CFS, the low tide (menstruation) leaves their boat stranded on the rocks, causing a crash. Using hormonal birth control is like building a sea wall that keeps the water level steady, preventing the boat from hitting the rocks.

This study gives patients the map they need to navigate their own tides and suggests a new way to help them stay afloat.

Technical Summary

1. Problem Statement

Long COVID and Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) disproportionately affect females (approx. 2:1 ratio), suggesting a potential modulation by sex hormones. While patients and preliminary surveys have reported that symptoms (particularly "crashes"—sudden, severe worsening after exertion) fluctuate with the menstrual cycle, existing research relies heavily on retrospective surveys subject to recall and recruitment bias. Furthermore, the specific impact of hormonal contraception on these conditions remains uninvestigated. The study aims to determine if symptom severity and crash frequency vary across menstrual cycle phases and if stabilizing ovarian hormones via contraception ameliorates symptoms.

2. Methodology

Study Design & Data Sources:

• Type: Retrospective analysis of prospectively collected, daily self-reported data.
• Disease Cohort: Users of the Visible app (designed for chronic conditions) with self-reported Long COVID, ME/CFS, or both. Data collected from Sept 7, 2022, to March 6, 2024.
• Population Cohort: Users of the Hertility app (menstrual tracker) serving as a general population comparison. Data collected from Sept 7, 2022, to Sept 28, 2025.
• Inclusion Criteria: Females/non-binary individuals aged 18–45 with regular menstrual cycles. Excluded were those pregnant, breastfeeding, or with irregular cycles.
• Sample Size: 948 users in the disease cohort (100% female-identifying or non-binary).

Data Collection & Variables:

• Symptoms: Daily tracking of 36 symptoms on a 4-point Likert scale (None to Severe).
• Crashes: Binary daily reporting of "crashes" (defined as sudden severe worsening affecting activity).
• Covariates: Menstrual phase (assigned via FIGO definitions), contraception type (None, Combined Hormonal, Progestin-only), disease type, age, gravidity, and time since app registration.
• Outcome Measures:
  1. Crash Incidence: Binary outcome (crash vs. no crash).
  2. Overall Symptom Score: Sum of daily symptom scores (0–108 range).
  3. Individual Symptoms: Fatigue, brain fog, and headache severity/presence.

Statistical Analysis:

• Software: R (v4.5.1).
• Models: Generalized Linear Mixed-Effects Models (GLMMs) with 'Individual' as a random effect to account for within-person correlation.
  • Crashes: Logistic regression (Binomial family, logit link) $\rightarrow$ Odds Ratios (OR).
  • Overall Symptom Score: Negative binomial regression (to handle overdispersed count data) $\rightarrow$ Incidence Rate Ratios (IRR).
  • Individual Symptoms: Cumulative Link Mixed Models (CLMM) for the disease cohort (ordinal data) and Logistic regression for the population cohort (binary data).
• Adjustments: Models controlled for age, disease type, contraception, gravidity, and quarters since joining. Multiple comparisons were adjusted using the Bonferroni method.

3. Key Contributions

• First Prospective App-Based Study: Unlike previous retrospective surveys, this study utilizes daily, real-time data collection, significantly reducing recall bias.
• Granular Cycle Analysis: It is the first study to quantify symptom changes across all menstrual phases (follicular, luteal, premenstrual, menses) rather than just the perimenstrual window.
• Contraception Impact: It is the first formal investigation into the effect of hormonal contraception on Long COVID and ME/CFS symptom severity.
• Population Comparison: By comparing disease cohorts with a general population cohort, the study distinguishes between disease-specific cyclical exacerbations and normal menstrual-related physiological fluctuations.

4. Key Results

Menstrual Cycle Phase Effects:

• Crash Frequency: Crashes were significantly more frequent during menstruation compared to all other phases.
  • The Late Luteal phase showed the greatest reduction in crash risk compared to menstruation (OR = 0.888, 95% CI: 0.838–0.941).
  • Other phases (Follicular, Early Luteal, Premenstrual) also showed significantly lower crash odds than menstruation.
• Symptom Severity: Overall symptom scores were highest during menstruation and lowest during the Early Luteal phase.
  • Compared to menstruation, the Early Luteal phase had the greatest reduction in symptom scores (IRR = 0.963, p < 0.001).
• Specific Symptoms: Fatigue, brain fog, and headaches were significantly more likely to be reported during menstruation compared to other phases in both disease and population cohorts. However, the magnitude of fluctuation was larger in the disease cohort.

Contraception Effects:

• Combined Hormonal Contraception (CHC): Users of CHC (pills/patch) had significantly fewer crashes (OR = 0.548, p = 0.008) and lower overall symptom scores (IRR = 0.827, p = 0.041) compared to non-users.
• Progestin-Only: No statistically significant reduction in crashes or symptom scores was observed compared to non-users.

Disease Type & Demographics:

• Users with both Long COVID and ME/CFS had the highest symptom burden and crash odds, followed by ME/CFS only, then Long COVID only.
• Age and gravidity were not significantly associated with crash incidence or overall symptom scores in the adjusted models.

Population Comparison:

• The general population cohort also reported higher fatigue, brain fog, and headache during menstruation, suggesting that some cyclical variation is a normal physiological phenomenon. However, the "crash" phenomenon is specific to the disease cohort.

5. Significance and Implications

• Clinical Management: The findings empower patients to anticipate cyclical symptom exacerbations (particularly during menstruation) and plan activities accordingly to avoid crashes.
• Therapeutic Potential: The association between Combined Hormonal Contraception (CHC) and reduced symptoms suggests a modulatory role of ovarian hormones (specifically the stabilization of estrogen/progesterone fluctuations). This highlights CHC as a potential, low-cost, non-invasive therapeutic strategy for managing Long COVID and ME/CFS, warranting further investigation in randomized controlled trials.
• Pathophysiology: The results support the hypothesis that dysregulated immune responses or neuroinflammation in these conditions are sensitive to hormonal fluctuations, reinforcing the need for sex-specific research in chronic post-viral syndromes.
• Limitations: The study is observational; residual confounding (e.g., health-seeking behavior differences between contraceptive users and non-users) cannot be excluded. Self-reported disease status and lack of direct hormone level measurements (relying on cycle phase proxies) are also limitations.

Conclusion:
This study provides robust evidence that ovarian hormone fluctuations significantly influence the severity of Long COVID and ME/CFS. Menstruation is a high-risk period for symptom exacerbation and crashes, while the use of combined hormonal contraception is associated with a clinically meaningful reduction in symptom burden.