Sex Differences in Long COVID, 2025, Shah et al.

[SNT Gatchaman]

Sex Differences in Long COVID
Dimpy P. Shah et 1226 al.

IMPORTANCE
A substantial number of individuals worldwide experience long COVID, or post-COVID condition. Other postviral and autoimmune conditions have a female predominance, but whether the same is true for long COVID, especially within different subgroups, is uncertain.

OBJECTIVE
To evaluate sex differences in the risk of developing long COVID among adults with SARS-CoV-2 infection.

DESIGN, SETTING, AND PARTICIPANTS
This cohort study used data from the National Institutes of Health (NIH) Researching COVID to Enhance Recovery (RECOVER)–Adult cohort, which consists of individuals enrolled in and prospectively followed up at 83 sites in 33 US states plus Washington, DC, and Puerto Rico. Data were examined from all participants enrolled between October 29, 2021, and July 5, 2024, who had a qualifying study visit 6 months or more after their initial SARS-CoV-2 infection.

EXPOSURE
Self-reported sex (male, female) assigned at birth.

MAIN OUTCOMES AND MEASURES
Development of long COVID, measured using a self-reported symptom-based questionnaire and scoring guideline at the first study visit that occurred at least 6 months after infection. Propensity score matching was used to estimate risk ratios (RRs) and risk differences (95% CIs). The full model included demographic and clinical characteristics and social determinants of health, and the reduced model included only age, race, and ethnicity.

RESULTS
Among 12 276 participants who had experienced SARS-CoV-2 infection (8969 [73%] female; mean [SD] age at infection, 46 [15] years), female sex was associated with higher risk of long COVID in the primary full (RR, 1.31; 95% CI, 1.06-1.62) and reduced (RR, 1.44; 95% CI, 1.17-1.77) models. This finding was observed across all age groups except 18 to 39 years (RR, 1.04; 95% CI, 0.72-1.49). Female sex was associated with significantly higher overall long COVID risk when the analysis was restricted to nonpregnant participants (RR, 1.50; 95%: CI, 1.27-1.77). Among participants aged 40 to 54 years, the risk ratio was 1.42 (95% CI, 0.99-2.03) in menopausal female participants and 1.45 (95% CI, 1.15-1.83) in nonmenopausal female participants compared with male participants.

CONCLUSIONS AND RELEVANCE
In this prospective cohort study of the NIH RECOVER-Adult cohort, female sex was associated with an increased risk of long COVID compared with male sex, and this association was age, pregnancy, and menopausal status dependent. These findings highlight the need to identify biological mechanisms contributing to sex specificity to facilitate risk stratification, targeted drug development, and improved management of long COVID.


Link | PDF (JAMA Network Open) [Open Access]

 

[SNT Gatchaman]

This prospective NIH RECOVER-Adult cohort study found that [female] sex was associated with an increased risk of long COVID compared with male sex and that the association was age, pregnancy, and menopausal status dependent. Understanding the mechanisms of sex differences can provide preventive and management strategies for not only long COVID but also other postviral illnesses.

For females the relative risks compared with males were—

18-39: 1.04
40-54: 1.48
≥55: 1.34

In the 40-54y group (excluding pregnant at follow-up) —

Non-menopausal: 1.45
Menopausal: 1.42

 

[Hutan]

They say that Long Covid risk is menopausal status dependent. But, 1.45 (non-menopause) versus 1.42 (menopause) in the 40-54 year female group (all relative to men)? That's basically the same. Maybe the hormones that change during menopause aren't affecting Long Covid onset risk.

Female sex risk overall was 1.31 relative to men, and they say that female sex risk was 1.50 in non-pregnant participants relative to men. That's quite a difference. There must have been a lot of pregnant people in the sample and/or their risk of Long Covid must have been drastically reduced. Maybe the number of pregnant people was very small.


Overall female risk relative men is not very high. A risk ratio of 1.31 means that females had 1.31 times the risk of getting Long Covid as men. I make that about 57% women versus 43% men. And in the 18 to 39 year age range, the risk of Long Covid was basically the same in men and women (perhaps particularly affected by the seeming low risk during pregnancy?)

These findings highlight the need to identify biological mechanisms contributing to sex specificity to facilitate risk stratification, targeted drug development, and improved management of long COVID.

I haven't read the paper yet, and I aim to because this is interesting. But I suspect what these findings highlight is that the definition of Long Covid is way too loose. I think it's important to stratify the sample according to what symptoms are persisting, and take into account any overt tissue damage, before anyone starts trying to make a story about biological mechanisms contributing to sex specificity.

They don't say in the abstract what percentage of people got "Long Covid" after a Covid-19 infection. That would give us a clue about how loose the definition is.

 

[Hutan]

The outcome was the presence of long COVID at the first visit 6 months or more after the index infection ascertained using a previously reported symptom-based scoring algorithm (ie, research index)10 that was updated in 2024.11 The symptoms contributing to the long COVID definition included postexertional malaise, fatigue, brain fog, dizziness, palpitations, loss of or change in smell or taste, thirst, chronic cough, chest pain, shortness of breath, and snoring or sleep apnea.

Participants with a research index of 11 or greater were considered long COVID positive; the rest were classified as long COVID indeterminate (not meeting classification criteria for long COVID but not necessarily asymptomatic).10,11 Participants positive for long COVID were assigned to long COVID subphenotypes based on Euclidean distance to the cluster centroid of each subphenotype.10,11

It's looking a bit more promising that I thought it might. They do seem to have stratified on symptoms.

 

[Jonathan Edwards]

Looks potentially very interesting. Even if there are major confounders the relatively narrow relative risk may be saying something important. I will need to look closer later.

 

[Hutan]

Symptoms and Long COVID by Sex

Symptom frequencies stratified by sex are provided in Table 2. A significantly higher proportion of females (1845 [21%]) compared with males (532 [16%]) had long COVID at the analysis visit (Table 3). The mean and median long COVID indices among long COVID–positive participants were slightly higher among females (mean [SD], 16.5 [4.7]; median, 15 [IQR, 13-20]) than among males (mean [SD], 15.9 [4.3]; median, 15 [IQR, 12-18]). The distribution of long COVID subphenotypes by sex is shown in Table 3. Symptom frequency by sex and long COVID subphenotype is shown in eFigure 4 in Supplement 1.

The participant characteristics are different for men and women e.g.

• 11% of the men are reported to have an immunocompromised condition, versus 4% of women.
• 30% of females were obese vs 21% of males.
• A staggering 40% of women reported a mental health disorder vs 26% of men.

But they say that they made adjustments in their model, so baseline differences probably don't matter so long as they had some coverage across each of the variables. I'm not sure how they got this baseline data - self-report?

The frequency of Long Covid symptoms in the Long covid group did not vary much between males and females.
From Table 2:
Symptom: Female %; Male %

Post-exertional malaise: F 88.6% ; M 88.3% (about 10% of the participants in the non-LC group (they call it LC indeterminate) are recorded as having PEM)
Fatigue: F 86.7%; M 83.2% (about 25% in non-LC women)
Dizziness: F 66.2%; M 66.5%

 

[hotblack]

There must have been a lot of pregnant people in the sample and/or their risk of Long Covid must have been drastically reduced. Maybe the number of pregnant people was very small.

Could it just be the effect of age here or did they account for that? Presumably those who were pregnant were younger and the difference looks similar to the age ratios?

 

[Hutan]

When 1755 participants (20%) who reported being pregnant between the index date and the study visit were excluded, female sex was associated with significantly higher risk of long COVID (RR, 1.50; 95%, CI, 1.27-1.77).

When comorbidities were added to the propensity score model, propensity score matching still remained balanced between males and females on all covariates (eFigure 2 inSupplement 1). However, there was no longer an association between sex and long COVID (RR, 1.07; 95% CI, 0.89-1.30).

Menopause and long COVID may also share some overlapping symptoms.

(So potentially inherently lower risk of LC but higher risk of menopause symptoms that are labelled as LC?)

A review of the literature17,21,44-47 combined with our data suggests that differences in hormonal levels may partially explain the higher prevalence of long COVID in females younger than 55 years. An attenuation in the risk ratio of long COVID for females aged 18 to 39 years may be explained by most pregnant individuals belonging to this age category.

 

[Hutan]

Could it just be the effect of age here or did they account for that? Presumably those who were pregnant were younger and the difference looks similar to the age ratios?

On the surface at least, they look to have accounted for things pretty well. But a lot of the information is in supplementary reports - there are annoying gaps in the report. The analysis around pregnancy is one.

An attenuation in the risk ratio of long COVID for females aged 18 to 39 years may be explained by most pregnant individuals belonging to this age category. In the sensitivity analysis excluding participants who were pregnant at any time between the index and study visit, female sex was still associated with an increased risk of long COVID.

I don't think those sentences make sense. The risk appears to have been lower in pregnant people, although I didn't see a risk ratio with confidence intervals reported. So, of course removing pregnant people from the analysis didn't lower the risk of long Covid in the remaining females.

Additionally, in our sensitivity analyses balanced on comorbidities, the RR for long COVID was attenuated. Many of these comorbidities (including myalgic encephalomyelitis/chronic fatigue syndrome [ME/CFS] and postural orthostatic tachycardia syndrome) are known to be associated both with female sex and with long COVID.17,18 Therefore, some of these comorbidities likely are mediators in the causal pathway and may partially explain the observed sex differences in the risk of long COVID.18

 

[Amw66]

They say that Long Covid risk is menopausal status dependent. But, 1.45 (non-menopause) versus 1.42 (menopause) in the 40-54 year female group (all relative to men)? That's basically the same. Maybe the hormones that change during menopause aren't affecting Long Covid onset risk.

Female sex risk overall was 1.31 relative to men, and they say that female sex risk was 1.50 in non-pregnant participants relative to men. That's quite a difference. There must have been a lot of pregnant people in the sample and/or their risk of Long Covid must have been drastically reduced. Maybe the number of pregnant people was very small.


Overall female risk relative men is not very high. A risk ratio of 1.31 means that females had 1.31 times the risk of getting Long Covid as men. I make that about 57% women versus 43% men. And in the 18 to 39 year age range, the risk of Long Covid was basically the same in men and women (perhaps particularly affected by the seeming low risk during pregnancy?)




I haven't read the paper yet, and I aim to because this is interesting. But I suspect what these findings highlight is that the definition of Long Covid is way too loose. I think it's important to stratify the sample according to what symptoms are persisting, and take into account any overt tissue damage, before anyone starts trying to make a story about biological mechanisms contributing to sex specificity.

They don't say in the abstract what percentage of people got "Long Covid" after a Covid-19 infection. That would give us a clue about how loose the definition is.

"Maybe the hormones that change during menopause aren't affecting Long Covid onset risk."

a hint at mechanisms? Estrogen and it's pathways do a lot of signalling and are involved in many areas . Ratios relative to other molecules seem to be key rather than absolute values.

Perhaps an aside but there seems to be a bit of recognition that vaccinations have played merry havoc with menstrual cycles over all ages groups.

Have hormone levels been considered ?

 

[Stuart79]

Is this good evidence that Long Covid is not a b-cell autoimmune disease given that it does not skew +-80% female?

 

[Andy]

PEM was defined as "Post-exertional malaise (Symptoms worse after even minor physical or mental effort)".

 

[Hutan]

In eTable 1, there are the results for the acute and crossover participants only. I think they are reported in the text too. These people were in the study with 30 days of their Covid-19 infection, so there is less room for self-selection when symptoms lingered. The risk ratio is a bit higher, 1.58.
The estimated proportion of LC (roughly at 6 months) is 11.4% for females and 7.2% for males.

eFigure 4 gives the symptom frequencies for 5 clusters, including cluster 3 which they call PASC. It has a lot of PEM, brain fog, post-exertional soreness and fatigue. Unfortunately, there is no report of the risk ratios by sex for the clusters. Cluster 5 seems to be of people who ticked pretty much every box.

There isn't that much in the supplementary material. There's quite a lot going on under the bonnet in this study I think, although it feels as though the researchers knew what they were doing.