For people already living with asthma or chronic obstructive pulmonary disease, recovering from COVID-19 can be difficult to separate from the respiratory symptoms they experienced before infection. A new Swedish study shows how substantial that overlap may be, while also identifying characteristics associated with persistent symptoms after infection.
Among 494 adults with asthma or COPD who reported confirmed COVID-19, 289, or 59%, said they experienced persistent, new or worsening symptoms for more than three months. The proportion was 55% among participants with asthma and 70% among those with COPD.
The strongest association in the adjusted analysis was with the severity of the original infection. Participants who described their acute COVID-19 as severe had 8.91 times the odds of reporting post-COVID symptoms compared with those whose infection was mild. Obesity, chronic pain and anxiety or depression were also associated with higher odds.
The findings, published in npj Primary Care Respiratory Medicine, do not establish that these factors cause post-COVID illness. The study was cross-sectional and relied substantially on self-report. But it provides a focused look at a population in which persistent breathlessness, fatigue and cough can be particularly difficult to interpret clinically.
Why chronic lung disease complicates the long-COVID picture
Asthma and COPD are common obstructive lung diseases, and respiratory infections can worsen both conditions. Post-COVID illness adds another layer because its symptoms can involve respiratory, neurological and systemic systems. Fatigue, breathlessness, cognitive problems and musculoskeletal symptoms are among the manifestations reported in previous research.
That creates an important measurement problem. A person with COPD may already experience breathlessness before SARS-CoV-2 infection. A person with asthma may have intermittent cough or respiratory limitation. Researchers therefore need to distinguish symptoms that simply continue from the underlying disease from symptoms that are new or have become worse after COVID-19.
The Swedish team addressed this by classifying participants as having post-COVID symptoms when they reported persistent, new or worsening symptoms lasting more than three months after infection. The definition captures prolonged change from a person’s previous state, although it does not fully reproduce every criterion used in formal clinical definitions of post-COVID condition.
The study drew on a long-running Swedish respiratory cohort
The researchers used data from the Swedish PRAXIS study. Participants had physician-diagnosed asthma or COPD recorded in medical records and had originally been recruited from primary and secondary care across eight regions in central Sweden.
In spring 2022, follow-up questionnaires were sent to surviving participants with known addresses. The wider responding cohort included 1,513 people with COPD and 1,787 with asthma. For the present analysis, the researchers included 2,328 participants with the information needed for their study. Of these, 44% had COPD and 56% had asthma.
A COVID-19 infection counted as confirmed when a participant reported a positive PCR, antigen or antibody test, or a diagnosis made by a doctor or nurse. People who thought they had been infected but had not been tested were grouped with those without confirmed COVID-19.
Overall, 22% of the 2,328 participants reported confirmed infection. Most first infections occurred from January 2022 onward, accounting for 55% of confirmed cases. Another 23% reported a first infection between March 2020 and February 2021, 12% between July and December 2021, and 11% between March and June 2021.
The researchers analysed associations using logistic regression, adjusting primarily for age and respiratory diagnosis. They also used a Benjamini-Hochberg false discovery rate procedure as a sensitivity analysis to reduce the risk that statistically significant findings arose simply because many comparisons were tested.
Nearly six in ten infected participants reported prolonged symptoms
Among the 494 participants with confirmed COVID-19, 289 reported symptoms lasting more than three months, producing the headline prevalence of 59%.
The symptom burden was not limited to a single complaint. A total of 147 participants reported one or two post-COVID symptoms, while 142 reported three or more. Fatigue, dyspnoea and cough or rhinitis were the most common symptoms in both the asthma and COPD groups.
The raw difference between diagnoses was sizeable: 55% of participants with asthma reported post-COVID symptoms compared with 70% of those with COPD. Yet COPD itself was not significantly associated with post-COVID symptoms compared with asthma after adjustment for age. That distinction matters because COPD is more common at older ages, making a simple comparison of percentages potentially misleading.
The study also found that asthma was associated with higher odds of having reported confirmed COVID-19 in the first place. Compared with COPD, the adjusted odds ratio was 1.52, with a 95% confidence interval from 1.18 to 1.97. The authors suggest that pandemic-era testing and isolation patterns may partly explain this finding, particularly because older people with COPD may have been more likely to avoid exposure during earlier waves.
Severity of the original infection stood out
The clearest statistical signal concerned acute disease severity. Participants reporting severe rather than mild acute COVID-19 had an adjusted odds ratio of 8.91 for later post-COVID symptoms. The 95% confidence interval ranged from 3.49 to 22.8.
That interval is wide, indicating uncertainty around the precise size of the association, but even its lower bound represents a substantial difference in odds. Hospitalisation was also associated with persistent symptoms in the main analysis, although that result did not remain statistically significant after the researchers corrected for multiple comparisons.
People who had not sought medical care during the acute infection showed lower odds of post-COVID symptoms, and that association survived the multiple-comparison sensitivity analysis. Together, these findings point in the same general direction: more serious acute illness was linked to a greater likelihood of prolonged symptoms.
Obesity and chronic pain remained important after stricter testing
Obesity, defined as a body mass index of at least 30, was associated with more than twice the odds of post-COVID symptoms. The adjusted odds ratio was 2.24, with a 95% confidence interval from 1.44 to 3.47.
Chronic pain produced an even larger association. Participants reporting chronic pain had 3.61 times the adjusted odds of post-COVID symptoms, with a 95% confidence interval from 1.95 to 6.69. Anxiety or depression was associated with an adjusted odds ratio of 1.79, with a 95% confidence interval from 1.13 to 2.80.
Importantly, severe acute infection, obesity, chronic pain and anxiety or depression remained associated with post-COVID symptoms after the false discovery rate correction. This makes those findings more robust within the study’s own statistical framework than associations that disappeared under the stricter sensitivity test.
Higher age and female sex were significant in the main adjusted analysis. For female sex, the adjusted odds ratio was 1.52, with a 95% confidence interval from 1.04 to 2.21. However, the sex association did not remain statistically significant after correction for multiple comparisons, so it warrants more cautious interpretation.
Earlier infections were linked to greater persistent symptom risk
The timing of infection also mattered. Participants who reported their first infection during March 2020 to February 2021, when the original Wuhan lineage predominated in Sweden, had higher odds of post-COVID symptoms than people infected later. First infection from January 2022 onward was associated with lower risk.
It is tempting to interpret this as a direct comparison of viral variants, but the study cannot support that conclusion. Calendar periods differ in far more than the dominant variant. Vaccination coverage increased, population immunity changed, testing practices evolved and clinical management improved over time.
The researchers themselves caution that the questionnaire did not explicitly link each participant’s post-COVID symptoms to the timing of the first reported infection. The time-period finding is therefore informative but not a clean test of whether one SARS-CoV-2 variant was more likely than another to produce long-term symptoms.
Why 59% should not be treated as a universal long-COVID rate
The 59% figure is striking, but its meaning depends heavily on the population and definition used. These participants already had asthma or COPD, and the outcome was based on self-reported persistent, new or worsening symptoms lasting more than three months.
The study did not establish whether the symptoms impaired everyday functioning, which is part of the World Health Organization’s definition of post-COVID condition. Nor could it always determine whether a symptom arose directly from COVID-19, reflected worsening underlying respiratory disease, or resulted from another health condition.
This helps explain why prevalence estimates differ so widely across studies. A registry study based only on formally coded diagnoses can produce a much lower rate because many people with persistent symptoms never receive a post-COVID diagnosis. A questionnaire that actively asks about symptoms may identify substantially more people.
The present result is therefore best understood as the proportion of this Swedish respiratory-disease sample who reported qualifying prolonged symptoms, not as an estimate that six in ten people with COVID-19 generally will develop long COVID.
The findings could help clinicians identify patients needing follow-up
For primary care, the study highlights a practical challenge. Patients with chronic airway disease may already have fatigue, cough and breathlessness, yet those same symptoms can worsen or persist after viral infection. Follow-up therefore needs to consider change from the patient’s previous baseline rather than the presence of a symptom alone.
The strong association with acute severity suggests that patients who experienced particularly severe infections may warrant closer monitoring. The associations with obesity, chronic pain and anxiety or depression may also help clinicians recognise groups in whom persistent symptoms are more common.
These associations should not be turned into deterministic predictions for individual patients. An odds ratio describes differences between groups and does not mean that a person with a particular characteristic will necessarily develop persistent symptoms. Nor does an association establish that changing the characteristic would eliminate the risk.
Several limitations constrain causal interpretation
The cross-sectional design is the most important limitation. Exposure characteristics and post-COVID symptoms were assessed from the same broad questionnaire framework, so the study can identify associations but cannot establish causal pathways.
Much of the information was self-reported. Participants were sometimes recalling infections that occurred up to two years earlier, creating scope for recall error. Confirmed infection was also defined through participants’ reports of testing or clinical diagnosis rather than through a single linked laboratory registry.
Only 494 participants were included in the confirmed COVID-19 analysis and 289 met the post-COVID symptom definition. That sample limited the number of confounding variables the researchers could include simultaneously. Residual confounding is therefore possible.
The study also lacked information on functional impairment and could not conclusively separate effects associated with pandemic period, viral variant, vaccination, clinical management and changing population immunity.
At the same time, the cohort has useful strengths. Participants were drawn from both primary and secondary care and originally selected through established respiratory-disease cohorts rather than recruited specifically because they had long COVID. That reduces some of the referral bias that can affect studies based only on specialist post-COVID clinics.
A focused signal in a clinically complicated population
The study does not settle how common long COVID is among everyone with asthma or COPD, nor does it prove why some people develop persistent symptoms. What it does show is that prolonged symptoms were common in this Swedish sample and were not distributed randomly across participants.
Acute severity produced the strongest association, while obesity, chronic pain and anxiety or depression remained significant even after a stricter correction for multiple testing. Those patterns can help guide future prospective research and may help clinicians decide which patients deserve particularly careful follow-up after respiratory viral infection.
For people with chronic airway disease, the central clinical question may be less whether a symptom exists than whether infection changed its severity, persistence or impact. This study provides evidence that such changes can remain substantial months after the acute infection has passed.
Source Information
Study: Risk factors of post-COVID-19 symptoms, a cross-sectional study in patients with asthma and chronic obstructive lung disease
Journal: npj Primary Care Respiratory Medicine, volume 36, article 64 (2026)
Study design: Cross-sectional analysis of the Swedish PRAXIS cohort, including 2,328 adults with asthma or COPD and a focused analysis of 494 participants reporting confirmed COVID-19.
Published: 29 September 2026
Source: https://doi.org/10.1038/s41533-026-00570-x
