Risk factors associated with chronic pain and fibromyalgia-like symptoms post-COVID hospitalization

Abstract

Post-COVID syndrome (PCS) may manifest as chronic pain and fibromyalgia-like symptoms, but the specific risk factors contributing to these outcomes remain poorly understood. We conducted a prospective cohort study using data from the ongoing “Post-Acute Sequelae of SARS-CoV-2 Infection (PASC)” study to identify predictors of chronic pain and fibromyalgia-related phenotypes among previously hospitalized COVID-19 patients. Participants were adults (n = 756, mean age 55, 47% female) hospitalized between March and August 2020 at the University of São Paulo Medical School. Clinical and demographic data from hospitalization, along with psychological, functional, and cognitive assessments at 3–11 months post-discharge (phase 1), were used to predict chronic pain, chronic widespread pain (CWP), fibromyalgia-like symptoms (FLS), and probable fibromyalgia (FM) at a 2–3 year follow-up (phase 2). Multivariate logistic regression analyses revealed that higher anxiety/depression severity (OR 1.35; 95% CI 1.09–1.69; p = 0.008) and reduced left handgrip strength (OR 0.95; 95% CI 0.92–0.98; p < 0.001) at phase 1 significantly predicted chronic pain at phase 2. For fibromyalgia-related phenotypes, greater left-side pain, insomnia severity, and weaker left handgrip strength were consistently associated with increased risk across all outcomes (CWP, FLS, FM), with statistically significant odds ratios ranging from 1.07 to 1.44. These findings highlight key modifiable risk factors – particularly psychological distress, sleep disruption, and muscle weakness – that may contribute to the development of chronic pain and fibromyalgia symptoms following COVID-19 hospitalization. Early identification and targeted intervention on these domains could improve long-term outcomes in PCS populations.

Keywords

Post-COVID syndrome chronic pain fibromyalgia risk factors

Introduction

With COVID-19 now classified as an endemic disease, it is critical to understand the post-acute sequelae of SARS CoV-2 infection, referred to as post-COVID syndrome (PCS). PCS – which has been reported in 10%–35% of patients and up to 85% of hospitalized patients – is often characterized by an unexplained persistence or emergence of new symptoms lasting over 12 weeks after infection.¹ Cornerstone features of PCS include fatigue (28%–85% prevalence), memory or concentration difficulties (24%), joint or muscle pain (45%–75%), sleep disturbances (20%), and depression (18%),^2–5 with higher rates of these symptoms observed in patients with more severe COVID-19.⁶ These symptoms bear a strong resemblance to the defining characteristics of chronic pain and fibromyalgia syndrome (FM).

Chronic pain is defined as any pain lasting more than 3 months, and chronic widespread pain (CWP) is a known symptom of FM.⁷ The overlap between PCS, chronic pain, and FM symptoms suggests some degree of interrelation between these disorders. In fact, patient surveys have revealed that 63.3% of post-acute COVID-19 patients experience chronic pain, and up to 30% have met modified 2011 American College of Rheumatology (ACR) criteria for FM.^8–10 However, the exact nature of these relationships remains largely unknown.

Given that infections such as the Epstein-Barr virus have already been linked to FM onset, it is plausible that abnormal inflammatory pathways altered by COVID-19 infection may induce CNS hypersensitivity and, consequently, FM symptoms.^11–13 Furthermore, the COVID-19 pandemic cultivated many negative environmental factors – such as mental stress and inactivity due to quarantine or hospital admission – that may trigger chronic pain and FM following COVID-19 infection.^14,15 No existing studies have connected various facets of behavioral, psychological, and clinical stressors related to COVID-19 infection to the subsequent development of chronic pain or fibromyalgia-like symptoms (FLS).

A limited number of cross-sectional studies on probable post-COVID FM have yielded inconsistent results; Akel et al.¹⁶ and Savin et al.¹⁷ identified female gender as a strong predictor, whereas Ursini et al.⁹ found male gender to be a significant predictor. Studies examining post-COVID chronic pain have revealed female sex, body mass index (BMI), COVID symptom load, pre-existing pain symptoms, and days hospitalized as risk factors.^18–22 These studies were primarily conducted through web-based questionnaires or phone interviews, which may limit reliability of results and involve self-selection of study participants. Furthermore, longitudinal relationships between post-COVID recovery and FM/FLS onset are not well characterized.

To address these existing gaps, this prospective cohort analysis aims to (1) characterize long-term chronic pain, CWP, FLS, and probable FM in a sample of post-hospitalized COVID-19 patients at two time points and (2) analyze potential sociodemographic, clinical, behavioral, and psychological risk/protective factors on the no-pain to probable FM continuum. This study will contribute to ongoing investigations of COVID-19 as a potential trigger for chronic pain and FM, providing new insights for post-COVID symptom management and risk-stratified treatment approaches.

Methods

Study design

This analysis consists of post-hospitalized COVID-19 survivors from the ongoing “Post-acute sequelae of SARS-CoV-2 infection (PASC)” study conducted at the General Hospital of the University of Sao Paulo Medical School. Eligible participants were 18 or older and hospitalized with confirmed COVID-19 infection for over 24 h between March and August 2020. All patients discharged during this period were invited to participate in this study. Phase 1 evaluation occurred 3–11 months after discharge, and phase 2 follow-up assessments took place 2–3 years after phase 1. Data collection was performed by a multidisciplinary team of clinical professionals via in-person assessments and teleconsultation questionnaires. Sociodemographic and clinical information, as well as a wide range of psychological, functional, physiological, and cognitive measures, were collected to provide a comprehensive patient health assessment. In this prospective cohort analysis, these measures were used to identify risk factors at phase 1 that are associated with the presence of post-COVID chronic pain, CWP, FLS, or probable FM in phase 2. The institutional review board of the Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo gave ethics approval for all protocol components for this study (approval numbers: 4.270.242, 4.502.334, 4.524.031, 4.302.745, and 4.391.560). Informed written consent were obtained from participants (or their legal guardians) prior to study procedures.

Experimental measures

Exposure measures included the Insomnia Severity Index, handgrip strength, WHO Disability Assessment Schedule (WHODAS – functioning across six domains), EQ-5D questionnaire (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression, self-reported health status VAS), and the Post-COVID Functional Status scale (post-COVID symptoms and functional limitations).^23,24 PCFS and EQ-5D scores, as ordinal categorizations, were treated as continuous variables to align with other measures of patient function.

Select measures were used to classify chronic pain and FM/FLS, as explained below and in eTable 1.

Classification criteria

Chronic pain

Chronic pain was classified based on the presence of any pain sites and associated visual analogue scale (VAS) scores (mild to severe) throughout the body, with the exception of pain sites associated with a previous injury unrelated to COVID-19 infection.

Chronic widespread pain

CWP was defined as pain present in at least four of the five Widespread Pain Regions – left upper, right upper, left lower, right lower, or axial – as described by the International Association for the Study of Pain (IASP).¹⁴

Fibromyalgia

To capture the wide range of FM diagnostic standards, two distinct sets of classification criteria were used to sort patients into having either no FM/FLS or probable FM/FLS in Phase 2. The presence of FM/FLS was stored separately as a binary variable for each definition.

Modified American College of Rheumatology Definition for FM. The modified 2010 ACR criteria for clinical studies (2011 ACR survey criteria), in combination with the 2016 ACR criteria, was used to classify FM. The Symptom Severity Score (SSS), ranging from 0 to 12, was calculated as the sum total of the severity of fatigue (0–3), waking unrefreshed (0–3), cognitive impairment (0–3), and presence of anxiety/depression, headaches, or abdomen pain (0–3). The Widespread Pain Index (WPI), ranging from 0 to 19, was calculated as the sum of reported pain areas matching the predefined body sites, as listed in Table 1. Using previously-validated criteria, patients were classified as having probable FM with either (1) SSS + WPI ≥ 13, (2) WPI ≥ 7 and SSI ≥ 5, or (3) WPI 3-6 and SSI ≥ 9.²⁵

Modified Definition for FLS. Patients were classified as having FLS if they reported pain that was more widespread than local or regional pain (pain in at least three of the five Widespread Pain Regions) and at least one complaint of moderate/severe fatigue, waking unrefreshed, cognitive impairments, headaches, or depression/anxiety.¹⁰

Table 1.

Patient characteristics at phase 1.

Characteristic	N	Overall (N = 756^a)	No pain (N = 226^a)	Pain (N = 530^a)	p Value^b
Female sex	756	356 (47%)	87 (38%)	269 (51%)	0.002
Age	733	55 (14)	53 (15)	56 (14)	0.054
BMI	605	32 (13)	31 (13)	32 (12)	0.10
Resting heart rate (BPM)	661	88 (256)	77 (16)	92 (299)	>0.9
Mobility (EQ-5D)	732	1.79 (1.05)	1.44 (0.90)	1.93 (1.08)	<0.001
Self-care (EQ-5D)	732	1.39 (0.84)	1.27 (0.81)	1.43 (0.85)	<0.001
Activities (EQ-5D	732	1.69 (1.05)	1.37 (0.85)	1.81 (1.09)	<0.001
Pain/discomfort (EQ-5D)	732	2.26 (1.15)	1.51 (0.94)	2.55 (1.09)	<0.001
Anxiety/depression (EQ-5D)	730	2.11 (1.19)	1.80 (1.06)	2.23 (1.21)	<0.001
Health status VAS score (EQ-5D)	731	73 (21)	80 (18)	70 (21)	<0.001
EPWORTH sleepiness scale	732	9.3 (5.7)	8.2 (5.4)	9.7 (5.8)	0.001
Insomnia severity index	732	7 (6)	5 (5)	8 (6)	<0.001
Pain (Right)	718	4.1 (3.5)	0.6 (1.9)	5.4 (3.0)	<0.001
Pain (Left)	712	3.87 (3.42)	0.69 (2.06)	5.04 (3.06)	<0.001
Handgrip strength (Right)	721	20 (13)	24 (13)	19 (12)	<0.001
Handgrip strength (Left)	718	18 (12)	22 (12)	17 (12)	<0.001
FACIT-Fatigue score	732	13 (8)	10 (6)	14 (8)	<0.001
WHO disability assessment schedule	732	21 (9)	17 (8)	22 (9)	<0.001
Length of stay (Days)	733	18 (18)	19 (18)	18 (17)	0.7
Post-COVID functional status	756	1.20 (1.12)	0.73 (0.97)	1.38 (1.13)	<0.001
Chronic pain at phase 2	341	147 (43%)	30 (33%)	117 (47%)	0.029
Chronic widespread pain at phase 2	308	24 (7.8%)	3 (3.4%)	21 (9.5%)	0.074
Fibromyalgia-like symptoms at phase 2	318	38 (12%)	5 (5.7%)	33 (14%)	0.036
Probable fibromyalgia at phase 2	318	14 (4.4%)	1 (1.1%)	13 (5.6%)	0.12

Numbers may not sum to totals due to the exclusion of missing data.

n (%); Mean (SD).

Wilcoxon rank sum test for numerical variables; Fisher’s exact test for categorical variables.

We assumed that the duration of symptoms reported by patients, as measured in assessments, reflects a sustained period of discomfort. Additionally, comorbid disorders at phase 2 were disregarded as a diagnosis of FM is valid irrespective of other diagnoses.

The FACIT-F scale, which ranges from 0 to 52, was used to characterize fatigue level, with lower scores indicating more severe fatigue. Scores under 43 are considered clinically relevant, and existing literature supports using scores below 30 to define a severe fatigue population.²⁶ To account for missing data, cognitive status was determined by both the Patient Memory Complaints Scale and total score from the Mini-Mental State Examination, with the more severe of the two used in the final diagnostic criteria. The EPWORTH Sleepiness Scale was used as a proxy for waking unrefreshed, a common symptom of FM. Depression and anxiety were determined using the Hospital Anxiety and Depression Scale. Headache and abdominal pain were determined through recorded pain site locations. Fatigue, cognitive function, and daytime sleepiness scores were re-coded into a four-point scale, ranging from 0 (no symptoms) to 3 (severe symptoms) based on standard assessment interpretations. Depression, anxiety, headache, and abdominal pain were converted to binary variables for the SSS calculation.

Study sample

A flow diagram of study participants is shown in Figure 1. Participants with no pain-related assessments at phase 2 (n = 415) were excluded from the sample of 756 COVID-19 survivors. Chronic pain status at phase 2 was classified for the remaining 341 participants, representing both pain persistence and development.

Figure 1.

Flow diagram of study sample.

For the CWP/FM/FLS portion of the study, patients with any rheumatological disease at Phase 1 were excluded from both chronic pain and non-chronic pain groups at phase 2 (n = 10 and n = 5, respectively) due to the high likelihood that they were already taking measures to manage their condition. Patients with CWP at phase 1 were removed in CWP comparisons, and patients with FLS at phase 1 were removed from FM/FLS comparisons because many measures relevant to FM classification were not available at phase 1. Classifications of CWP, FLS, and FM were not mutually exclusive. These classifications formed a continuum between patients with chronic pain, CWP, FLS, and finally probable FM.

Statistical analysis

Phase 1 demographics, as well as a range of clinical and functional characteristics, were characterized for the sample of participants. Four comparisons were made based on phase 2 outcome classifications: (1) chronic pain versus no chronic pain (2) CWP versus no CWP (n = 308) (3) FM versus no FM (n = 318) and (4) FLS versus no FLS (n = 318). Comparisons of relative frequencies or means between these groups were conducted using Fisher’s exact test and Wilcoxon rank sum test, respectively. The specific test used is mentioned in the footnote of Table 1.

Bivariate and adjusted multivariate logistic regressions were used to assess predictive relationships between various measures and post-COVID outcome at phase 2. Sex, age at hospital admission, and resting heart rate – an indicator of general patient health – were established as fixed covariates in these analyses. Missing data were removed in individual analyses. Results are presented as odds ratios with 95% confidence intervals, and p values <0.05 were considered statistically significant. All statistical analyses were conducted using Python and R.

Results

Baseline results

Clinical characteristics of this sample were detailed in a prior analysis by Battistella et al.²⁷ The mean age at hospital admission was 55 years and the average length of hospital stay was 18 days. Of the 801 post-hospitalized COVID-19 patients initially included in the sample, 62% (497) of the patients were admitted to intensive care. During their hospitalization, 41.6% (333) were intubated.²⁷ Linear regression models indicated that intubation status – which can be regarded as a measure of COVID-19 severity – had no significant impact on VAS pain scores in phase 1. Furthermore, participants’ pain VAS severity classifications did not show a significant difference when grouped by the time between hospital discharge and phase 1 assessment.

Chronic pain

Chronic pain was identified in 70.1% of (530 of 756) patients at phase 1, an average of 6.8 months post-hospitalization. Females made up 51% of chronic pain patients, as compared to 38% of the patients with no chronic pain (p = 0.002). Patients with chronic pain at phase 1 had more moderate to severe problems across all five dimensions of the EQ-5D assessment (mobility, self-care, usual activities, pain/discomfort; p < 0.001; Table 1). They also had more problems with sleepiness and insomnia (EPWORTH: 9.7 vs 8.2, p = 0.001; ISI: 8 vs 5, p < 0.001), weaker handgrip strength (R: 19 vs 24, p < 0.001; L: 17 vs 22, p < 0.001), worse WHODAS performance (p < 0.001), and moderate to severe post-COVID functional limitations (p < 0.001). Similar associations were found in patients with chronic pain at phase 2 (an average of 2.7 years follow up from phase 1), with the addition of higher BMI (33 vs 31, p = 0.029) and higher systolic blood pressure (136 vs 136, p = 0.012; eTable 4).

Of the 530 patients with chronic pain at phase 1, 47% (117) had chronic pain at phase 2, compared to 33% of 226 in the no pain group (p = 0.029; Table 1). Chronic pain was classified in 43% (147 of 341) of patients at phase 2. Pain sites at phase 2 appeared to be more diffuse compared to phase 1, although most commonly reported pain sites remained to be the right knee (118 phase 1 counts, 51 phase 2 counts), left knee (106 phase 1 counts, 48 phase 2 counts), and lower back (90 phase 1 counts, 33 phase 2 counts; eFigure 1).

Both bivariate and adjusted multivariate analyses revealed many associations between exposure measures and chronic pain development (Table 2, eTable 3). To further examine these relationships, adjusted multivariate logistic regression models including a combination of exposure measures showed that severity of anxiety/depression (Odds Ratio (OR) 1.35; 95% Confidence Interval (CI) 1.09, 1.69; p = 0.008) and weaker left handgrip strength (OR 0.95; 95% CI 0.92, .98; p = 0.00) at phase 1 are statistically significant predictors for the presence of chronic pain at phase 2 (Table 3). It is important to mention that both weaker right and left handgrip strength were associated with the presence of chronic pain at phase 2, as presented in Table 2. However, we included only left handgrip strength because it improved the goodness-of-fit of the final model. Therefore, handedness dominance was not an important factor in these results. Pain severity did not appear to have a significant effect.

Table 2.

Bivariate analyses by chronic pain, CWP, FLS, and FM at phase 2.

Predictor	Chronic pain		Chronic widespread pain		Fibromyalgia-like symptoms		Probable fibromyalgia
Predictor	OR (95% CI)	p Value	OR (95% CI)	p Value	OR (95% CI)	p Value	OR (95% CI)	p Value
Male sex	0.6 (0.39, 0.92)	0.021	0.29 (0.11, 0.69)	0.008	0.48 (0.24, 0.96)	0.039	0.43 (0.13, 1.28)	0.141
Age	1.01 (1, 1.03)	0.118	1.03 (0.99, 1.06)	0.124	1.01 (0.99, 1.04)	0.398	0.98 (0.95, 1.02)	0.449
BMI	1.01 (1, 1.04)	0.152	1.02 (1, 1.05)	0.054	1.02 (0.99, 1.04)	0.170	1.03 (1, 1.06)	0.042
Resting heart rate (BPM)	1 (1, 1.01)	0.610	0.98 (0.95, 1)	0.330	0.98 (0.95, 1)	0.144	0.99 (0.95, 1)	0.772
Mobility (EQ-5D)	1.26 (1, 1.58)	0.049	1.44 (0.95, 2.09)	0.066	1.25 (0.88, 1.74)	0.195	1.77 (1.09, 2.78)	0.015
Self-care (EQ-5D)	1.19 (0.86, 1.65)	0.293	1.86 (1.13, 2.94)	0.008	1.39 (0.83, 2.16)	0.165	2.06 (1.14, 3.46)	0.008
Usual activities (EQ-5D)	1.24 (0.99, 1.56)	0.064	1.72 (1.2, 2.43)	0.002	1.22 (0.85, 1.68)	0.256	1.91 (1.22, 2.93)	0.003
Pain/discomfort (EQ-5D)	1.29 (1.06, 1.57)	0.011	1.48 (1.02, 2.15)	0.041	1.42 (1.04, 1.94)	0.025	1.86 (1.14, 3.13)	0.015
Anxiety/depression (EQ-5D)	1.3 (1.08, 1.57)	0.007	2.21 (1.53, 3.29)	0.000	1.77 (1.32, 2.4)	0.000	3.42 (1.98, 6.77)	0.000
Health status VAS score (EQ-5D)	0.98 (0.97, 0.99)	0.000	0.96 (0.95, 0.98)	0.000	0.98 (0.96, 0.99)	0.007	0.97 (0.95, 0.99)	0.004
EPWORTH sleepiness scale	1.03 (0.99, 1.07)	0.105	1.09 (1.01, 1.18)	0.019	1.11 (1.05, 1.18)	0.001	1.19 (1.08, 1.32)	0.001
Insomnia severity index	1.04 (1, 1.08)	0.037	1.15 (1.08, 1.23)	0.000	1.07 (1.02, 1.13)	0.009	1.13 (1.04, 1.23)	0.004
Pain (Right)	1.06 (0.99, 1.13)	0.078	1.24 (1.08, 1.45)	0.004	1.15 (1.03, 1.29)	0.013	1.19 (1, 1.45)	0.063
Pain (Left)	1.1 (1.03, 1.18)	0.005	1.27 (1.1, 1.48)	0.002	1.21 (1.08, 1.36)	0.001	1.41 (1.15, 1.8)	0.002
Handgrip strength (Right)	0.96 (0.94, 0.98)	0.000	0.92 (0.88, 0.96)	0.001	0.94 (0.91, 0.97)	0.001	0.94 (0.88, 0.99)	0.026
Handgrip strength (Left)	0.95 (0.93, 0.97)	0.000	0.91 (0.86, 0.95)	0.000	0.93 (0.89, 0.96)	0.000	0.93 (0.86, 0.98)	0.018
FACIT-Fatigue score	1.04 (1.01, 1.07)	0.018	1.11 (1.06, 1.17)	0.000	1.09 (1.04, 1.14)	0.000	1.17 (1.09, 1.26)	0.000
WHO disability assessment schedule	1.04 (1.01, 1.07)	0.008	1.1 (1.05, 1.15)	0.000	1.07 (1.03, 1.11)	0.000	1.14 (1.08, 1.21)	0.000
Length of stay (Days)	0.99 (0.97, 1)	0.088	0.96 (0.91, 1)	0.118	0.96 (0.92, 0.99)	0.045	0.9 (0.79, 0.98)	0.045
Post-COVID functional status	1.18 (0.96, 1.45)	0.113	2.12 (1.46, 3.13)	0.000	1.65 (1.21, 2.26)	0.002	2.37 (1.48, 3.9)	0.000

CI: confidence interval; OR: odds ratio. Boldface values represent statistically significant results.

Table 3.

Adjusted multivariate logistic regression models for the presence of chronic pain at phase 2.

Predictor (phase 1)	Chronic pain (1)		Chronic widespread pain (1)		Chronic pain (2)		Chronic widespread pain (2)
Predictor (phase 1)	OR (95% CI)	p Value	OR (95% CI)	p Value	OR (95% CI)	p Value	OR (95% CI)	p Value
Male sex	1.53 (0.84, 2.83)	0.172	1.31 (0.39, 4.34)	0.659	1.57 (0.85, 2.92)	0.151	1.57 (0.43, 5.66)	0.484
Age at admission	1.01 (0.99, 1.03)	0.183	1.03 (0.99, 1.07)	0.134	1.01 (0.99, 1.03)	0.173	1.04 (1, 1.08)	0.081
Resting heart rate	1 (1, 1.01)	0.545	0.99 (0.95, 1)	0.511	1 (1, 1.01)	0.538	0.99 (0.95, 1)	0.609
Anxiety/depression (EQ-5D)	1.35 (1.09, 1.69)	0.008	2.21 (1.47, 3.47)	0.000	1.3 (1.02, 1.65)	0.032	1.73 (1.12, 2.77)	0.016
Handgrip strength (Left)	0.95 (0.92, 0.98)	0.000	0.91 (0.85, 0.97)	0.005	0.95 (0.92, 0.98	0.000	0.91 (0.84, 0.97)	0.007
Insomnia severity index					1.02 (0.97, 1.07)	0.398	1.14 (1.05, 1.24)	0.002

CI: confidence interval; OR: odds ratio. Boldface values represent statistically significant results.

Chronic widespread pain, fibromyalgia-like symptoms, and fibromyalgia

CWP was found in 5% (42 of 756) of patients at phase 1 and 7.7% (24 of 308) of patients in phase 2. FLS were found in 12% (38 of 318) of patients in phase 2. Further, 4.4% (14 of 318) of patients in phase 2 were classified as having probable FM. Phase 1 characteristic comparisons based on the separate classifications at phase 2 are summarized in eTable 2. Results of bivariate and multivariate analyses are displayed in Table 2 and eTable 3.

Similar associations were observed when using CWP as the outcome variable in the chronic pain model. However, including insomnia as an exposure measure for CWP enhanced the model’s performance but slightly attenuated the effect size of anxiety/depression (Anxiety/Depression: OR 1.73; 95% CI 1.12, 2.77; p = 0.016. Handgrip Strength (Left): OR 0.91; 0.84, 0.97; Insomnia Severity Index: OR 1.14; 1.05, 1.24; p = 0.002; Table 3). On the other hand, insomnia was not a significant predictor for chronic pain in general.

In adjusted multivariate models for the presence of CWP/FLS/FM at phase 2 with a combination of exposures, more severe left side pain (CWP: OR 1.21; 95% CI 1.01, 1.5; p = 0.049. FLS: OR 1.16; 1.01, 1.33; p = 0.037. FM: OR 1.44; 1.11, 2.04; p = 0.016), insomnia (CWP: OR 1.17; 1.08, 1.27; p = 0.000. FLS: OR 1.07; 1.01, 1.15; p = 0.026. FM: OR 1.12; 1.02, 1.24; p = 0.022), and weaker left side handgrip strength (CWP: OR 0.91; 0.84, 0.97; p = 0.010. FLS: OR 0.93; 0.88, 0.98; p = 0.005. FM: OR 0.92; 0.83, 0.99; p = 0.042) at phase 1 were statistically significant predictors (Table 4). Similarly to chronic pain, both weaker right and left handgrip strength were associated with the presence of CWP/FLS/FM at phase 2, as presented in Table 2. However, we included only left handgrip strength because it improved the goodness-of-fit of the final model. Therefore, handedness dominance was not an important factor in these results. Neither the EQ-5D measurement of pain/discomfort nor the post-COVID functional status presented a significant effect.

Table 4.

Adjusted multivariate logistic regression model for CWP/FM/FLS at phase 2.

Predictor (phase 1)	Chronic widespread pain (3)		Fibromyalgia-like symptoms		Fibromyalgia
Predictor (phase 1)	OR (95% CI)	p Value	OR (95% CI)	p Value	OR (95% CI)	p Value
Male sex	1.78 (0.48, 6.53)	0.380	1.94 (0.73, 5.24)	0.186	2.94 (0.61, 14.71)	0.177
Age at admission	1.01 (0.97, 1.05)	0.531	1 (0.97, 1.03)	0.944	0.97 (0.92, 1.02)	0.224
Resting heart rate	0.98 (0.93, 1)	0.316	0.98 (0.94, 1)	0.158	0.98 (0.93, 1)	0.543
VAS pain (Left)	1.21 (1.01, 1.5)	0.049	1.16 (1.01, 1.33)	0.037	1.44 (1.11, 2.04)	0.016
Insomnia severity index	1.17 (1.08, 1.27)	0.000	1.07 (1.01, 1.15)	0.026	1.12 (1.02, 1.24)	0.022
Handgrip strength (Left)	0.91 (0.84, 0.97)	0.010	0.93 (0.88, 0.98)	0.005	0.92 (0.83, 0.99)	0.042

Adjusted for Sex, Age at Admission, and Resting Heart Rate. CI: confidence interval; OR: odds ratio. Boldface values represent statistically significant results.

Interestingly, between phase 1 and phase 2, pain severity improved in patients with no previous pain, which was not observed in patients with chronic pain at phase 2 (eTable 5). Handgrip strength improved in both no pain and chronic pain groups, whereas ISI scores increased in severity in both groups.

Discussion

In this study, we tracked relevant exposure measures extending many years post-hospitalization to predict symptoms resembling chronic pain, CWP, FLS, and FM in a sample of patients from Brazil – a country significantly impacted by COVID-19. A key strength of our analysis is the 2.7-year follow-up, which allows for sufficient time for chronic pain and fibromyalgia-related networks to develop. We identified weaker handgrip strength and more severe anxiety/depression to be risk factors for long-term chronic pain as well as chronic widespread pain. However, moving toward the FM state, sleep-related factors became more significant than anxiety or depression.

Pain severity and insomnia scores did not improve from phase 1 in patients with chronic pain at phase 2, suggesting potential chronification rather than recovery. Accordingly, 80% of the patients with chronic pain at phase 2 had chronic pain at phase 1. The prevalence of chronic pain at phase 1 (70.1%) and phase 2 (43%) is comparable to findings in similar studies of hospitalized patients with PCS and approximately double that of the general population (20.9%), potentially pointing to the lasting effects of COVID-19 infection.²⁸

Unlike other studies, FLS was evaluated on a continuum ranging from chronic pain to probable FM. Comparisons revealed the differential roles of anxiety/depression and insomnia for general chronic pain and chronic pain resembling FM. Our study expands on previous research by highlighting anxiety, depression, and weaker handgrip strength as significant long-term predictors for general chronic pain. These results are supported in non-COVID samples, where anxiety/depression have been associated with pain chronification.^28–30

While the intensity of pain at phase 1 was not linked to general chronic pain at phase 2, it was significantly associated with CWP/FM/FLS at phase 2. This suggests that factors underlying pain intensity in the acute phase of COVID-19 recovery may play a role in triggering characteristic symptoms of a fibromyalgia-like state. The significant effect of more intense pain, weaker handgrip strength, and insomnia as predictors of FM supports results found in non-COVID samples.^31–33 The lack of an effect of anxiety/depression for FLS may indicate that psychological effects may be more of consequence of experiencing FLS rather than a causative variable. Surprisingly, potential markers of inflammation such as BMI did not survive in our final models for chronic pain/FLS, despite being implicated as risk factors in other studies.¹⁸ An explanation for this result could be that pain-related PCS could arise from an interaction between inflammation and the identified risk factors of insomnia, and muscle weakness. Inflammation may lead to a more vulnerable neurological state and exacerbate the impact of these factors on triggering fibromyalgia-like symptoms. Other potential physiological pathways to explain the lack of association are changes in neuropeptides expression (e.g. CGRP blood levels), immune dysregulation (unbalanced lymphocyte populations and autoimmune reactions), or central sensitization (altered conditioned pain modulation test), which could account for these differential changes in pain sensitivity following COVID-19 hospitalization. Future studies should explore these biomarkers as potential mechanistic variables for post-COVID pain syndromes. Moreover, our findings further underscore the importance of controllable physical factors – such as strength training and sleep improvement interventions – in a post-COVID context.

Several mechanisms could explain the relationship between these post-COVID predictive variables and associated long-term pain syndromes. Handgrip strength is often considered a marker of muscle weakness and poor overall health status.³⁴ Performance may reflect the extent of COVID-19 infections’ impact in deteriorating muscle and joint strength throughout the body. Anxiety or depression, along with poor sleep quality, may lower patient’s pain threshold and lead to FLS.^35,36 The severity of these measures may be triggered or exacerbated by both COVID-19 and hospitalization. Given that FM may also lead to more severe anxiety and muscle weakness, further understanding is needed on the temporal interplay between these measures and symptom chronification.^30,37 Ultimately, aligning with existing theories, these syndromes may arise due to a persistent cycle of pain, functional limitations, and psychological stressors.¹⁵

Across all the models, the PCFS score – which measures the direct impact of COVID-19 on functional status – did not appear to be a significant predictor in any of the models despite being significant in bivariate analyses. Thus, it appears that CWP/FLS/FM may be affected by manageable risk factors after COVID-19 infection – such as handgrip strength – rather than the impacts of the infection themselves.

Clinical implications

Our findings have important public health implications for the management and/or prevention of chronic pain and FLS/FM upon a potential trigger such as COVID-19 infection. The strongest predictive variables – poor sleep quality and handgrip strength – can be targeted early on in addition to existing pain management techniques. Muscle strength and mass, as indicated by handgrip strength, can also be improved through home-based training and stretching exercises. Additionally, poor sleep quality, which is often overlooked in rehabilitation programs, may be critical to address to prevent further development of post-COVID symptoms. Early identification and management of these measures by healthcare providers or the patients themselves may be key to slowing the cycle of post-COVID related symptom chronification.

Limitations

Limitations of this analysis include the absence of a direct set of questions specific to the ACR FM diagnostic criteria. We aimed to control for this through establishing the continuum of PCS from chronic pain to probable FM. Additionally, most of the measures assessed at Phase 1 and Phase 2 could not be evaluated during hospital admission or immediately afterward due to logistical complications arising from the severity of the pandemic at the time. Additionally, the present study have potential risk of selection bias due to sample attrition, 415 subjects were excluded from consideration because of no pain assessment in phase 2 follow-up. However, it is unlikely, this subset of sample would modify the results, since the sample characteristics at phase 1 were homogeneous (Table 1). Further, this study was conducted on a sample of patients at one hospital in Brazil, which may limit the generalizability of our findings to moderate to severe COVID-19 patients. Future analyses could expand the sample size to patients affected by different severities of COVID-19 and employ more specific measures of FM at multiple timepoints after infection to validate our findings.

Conclusions

Our findings indicate that PCS has long-term impacts that resemble chronic pain and fibromyalgia symptoms at a higher rate than the general population. In our sample, significant risk factors for both regional and widespread chronic pain over an average of 2.7 years were identified to be anxiety/depression and weaker handgrip strength. Risk factors for FLS/FM include high baseline pain, insomnia, and weaker handgrip strength. Understanding these risks inform prevention strategies to reduce the risk of chronic pain associated with PCS.

Supplemental Material

sj-docx-1-mpx-10.1177_17448069251369069 – Supplemental material for Risk factors associated with chronic pain and fibromyalgia-like symptoms post-COVID hospitalization

Supplemental material, sj-docx-1-mpx-10.1177_17448069251369069 for Risk factors associated with chronic pain and fibromyalgia-like symptoms post-COVID hospitalization by Linamara Rizzo Battistella, Marta Imamura, Xinyi Christine Zhang, Kevin Pacheco-Barrios and Felipe Fregni in Molecular Pain

Footnotes

Acknowledgements

We are thankful for the support received from members of the HCFMUSP COVID-19 Study Group.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was partially supported by donations from the general public under the HC-COMVIDA crowdfunding scheme () with funds managed by the Fundação Faculdade de Medicina – USP.

ORCID iD

Kevin Pacheco-Barrios

Supplemental Material

Supplemental material for this article is available online.

References

Pavli

Theodoridou

Maltezou

. Post-COVID syndrome: incidence, clinical spectrum, and challenges for primary healthcare professionals. Arch Med Res 2021; 52(6): 575–581.

Fernandez-de-las-Peñas

Notarte

Macasaet

Velasco

Catahay

Ver

Chung

Valera-Calero

Navarro-Santana

. Persistence of post-COVID symptoms in the general population two years after SARS-CoV-2 infection: a systematic review and meta-analysis. J Infect 2024; 88(2): 77–88.

Kim

Bae

Chang

Kim

. Long COVID prevalence and impact on quality of life 2 years after acute COVID-19. Sci Rep 2023; 13(1): 11207.

Fitzpatrick

Cope

Bertolli

Sotoodehnia

West

Gentile

Unger

. Estimates of incidence and predictors of fatiguing illness after SARS-CoV-2 infection. Emerg Infect Dis 2024; 30(3): 539–547.

Calvache-Mateo

Navas-Otero

Heredia-Ciuró

Matín-Núñez

Torres-Sánchez

López-López

Valenza

. Post-COVID patients with new-onset chronic pain two years after infection: clinical and psychosocial profile. Pain Manag Nurs 2023; 24(6): 571–579.

Fernández-de-Las-Peñas

de-la-Llave-Rincón

Ortega-Santiago

Ambite-Quesada

Gómez-Mayordomo

Cuadrado

Arias-Navalón

Hernández-Barrera

Martín-Guerrero

Pellicer-Valero

Arendt-Nielsen

. Prevalence and risk factors of musculoskeletal pain symptoms as long-term post-COVID sequelae in hospitalized COVID-19 survivors: a multicenter study. Pain 2022; 163(9): e989–e996.

Yong

Mullins

Bhattacharyya

. Prevalence of chronic pain among adults in the United States. Pain 2022; 163(2): e328–e332.

Zis

Ioannou

Artemiadis

Christodoulou

Kalampokini

Hadjigeorgiou

. Prevalence and determinants of chronic pain post-COVID; cross-sectional study. J Clin Med 2022; 11(19): 5569.

Ursini

Ciaffi

Mancarella

Lisi

Brusi

Cavallari

D’Onghia

Mari

Borlandelli

Faranda Cordella

La Regina

Viola

Ruscitti

Miceli

De Giorgio

Baldini

Borghi

Gasbarrini

Iagnocco

Giacomelli

Faldini

Landini

Meliconi

. Fibromyalgia: a new facet of the post-COVID-19 syndrome spectrum? Results from a web-based survey. RMD Open 2021; 7(3): e001735.

10.

Wolfe

. The history of the idea of widespread pain and its relation to fibromyalgia. Scand J Pain 2020; 20(4): 647–650.

11.

Kocyigit

Akyol

. The relationship between COVID-19 and fibromyalgia syndrome: prevalence, pandemic effects, symptom mechanisms, and COVID-19 vaccines. Clin Rheumatol 2022; 41(10): 3245–3252.

12.

Ketenci

Zure

Akpınar

Soluk Özdemir

Balbaloğlu

Akaltun

Erden

Çağlıyan Türk

Korkmaz

Metin Ökmen

Altındağ

Soyupek

Yakşi

Sindel

Sezgin

Ustaömer

Kesiktaş

Dere

Güneş

Medin Ceylan

Sonel Tur

Evcik

. Pain types and risk factors in post-COVID-19. Turk J Phys Med Rehabil 2024; 70(1): 30–38.

13.

Gavrilova

Soprun

Lukashenko

Ryabkova

Fedotkina

Churilov

Shoenfeld

. New clinical phenotype of the post-Covid syndrome: fibromyalgia and joint hypermobility condition. Pathophysiology 2022; 29(1): 24–29.

14.

Dizner-Golab

Lisowska

Kosson

. Fibromyalgia – etiology, diagnosis and treatment including perioperative management in patients with fibromyalgia. Reumatologia 2023; 61(2): 137–148.

15.

Fialho

MFP

Brum

Oliveira

. Could the fibromyalgia syndrome be triggered or enhanced by COVID-19? Inflammopharmacology 2023; 31(2): 633–651.

16.

Akel

Almanasyeh

Abo Kobaa

Aljabali

Al-Abadleh

Alkhalaileh

Alwardat

Sarhan

Abu-Jeyyab

. A cross-sectional study of fibromyalgia and Post-acute COVID-19 Syndrome (PACS): could there be a relationship? Cureus 2023; 15(7): e42663.

17.

Savin

Rosenn

Tsur

Hen

Ehrenberg

Gendelman

Buskila

Halpert

Amital

. The possible onset of fibromyalgia following acute COVID-19 infection. PLoS One 2023; 18(2): e0281593.

18.

Ebbesen

Giordano

Valera-Calero

Hedegaard

Fernández-de-las-Peñas

Arendt-Nielsen

. Prevalence and risk factors of de novo widespread post-COVID pain in nonhospitalized COVID-19 survivors: a nationwide exploratory population-based survey. J Pain 2024; 25(1): 1–11.

19.

Moreira de Barros

da Silva

DIC

Barbosa

MLA

Soares

Alves

Miranda

Costa

PDLD

Nascimento Júnior

Módolo

NSP

. Chronic pain after hospital discharge on patients hospitalized for COVID-19: an observational study. Braz J Anesthesiol 2024; 74(4): 744457.

20.

Wangnamthip

Zinboonyahgoon

Rushatamukayanunt

Papaisarn

Pajina

Jitsinthunun

Promsin

Sirijatuphat

Fernández-de-Las-Peñas

Arendt-Nielsen

Ciampi de Andrade

. The incidence, characteristics, impact and risk factors of post-COVID chronic pain in Thailand: a single-center cross-sectional study. PLoS One 2024; 19(1): e0296700.

21.

Romeiser

Morley

Singh

. COVID-19 symptom load as a risk factor for chronic pain: a national cross-sectional study. PLoS One 2023; 18(6): e0287554.

22.

Fernández-de-las-Peñas

Pellicer-Valero

Martín-Guerrero

Hernández-Barrera

Arendt-Nielsen

. Investigating the fluctuating nature of post-COVID pain symptoms in previously hospitalized COVID-19 survivors: the LONG-COVID-EXP multicenter study. Pain Rep 2024; 9(3): e1153.

23.

Norén

Karlsson

Ohlsson-Nevo

Möller

Hermansson

. Psychometric evaluation of the WHODAS 2.0 and prevalence of disability in a Swedish general population. J Patient-Rep Outcomes 2023; 7: 36.

24.

Klok

Boon

GJAM

Barco

Endres

Geelhoed

JJM

Knauss

Rezek

Spruit

Vehreschild

Siegerink

. The post-COVID-19 functional status scale: a tool to measure functional status over time after COVID-19. Eur Respir J 2020; 56(1): 2001494.

25.

Wolfe

Clauw

Fitzcharles

M-A

Goldenberg

Häuser

Katz

Mease

Russell

Winfield

. Fibromyalgia criteria and severity scales for clinical and epidemiological studies: a modification of the ACR Preliminary Diagnostic Criteria for Fibromyalgia. J Rheumatol 2011; 38(6): 1113–1122.

26.

Eek

Ivanescu

Corredoira

Meyers

Cella

. Content validity and psychometric evaluation of the Functional Assessment of Chronic Illness Therapy-Fatigue scale in patients with chronic lymphocytic leukemia. J Patient Rep Outcomes 2021; 5(1): 27.

27.

Battistella

Imamura

De Pretto

Van Cauwenbergh

SKHAA

Delgado Ramos

Saemy Tome Uchiyama

Matheus

Kuhn

Amaral de Oliveira

Souza Naves

Rossetti Mirisola

de Quadros Ribeiro

Tadeu Sugawara

Cantarino

Andrade Santos Antunes Cavalca

Pagano

Valentim Marques

Mendes da Silva

Pereira Gomes

Fregni

. Long-term functioning status of COVID-19 survivors: a prospective observational evaluation of a cohort of patients surviving hospitalisation. BMJ Open 2022; 12(7): e057246.

28.

Macfarlane

. Generalized pain, fibromyalgia and regional pain: an epidemiological view. Best Pract Res Clin Rheumatol 1999; 13(3): 403–414.

29.

Magni

Moreschi

Rigatti-Luchini

Merskey

. Prospective study on the relationship between depressive symptoms and chronic musculoskeletal pain. Pain 1994; 56(3): 289–297.

30.

MacFarlane

Thomas

Papageorgiou

Schollum

Croft

Silman

. The natural history of chronic pain in the community: a better prognosis than in the clinic? J Rheumatol 1996; 23(9): 1617–1620.

31.

Markkula

Kalso

Kaprio

. Predictors of fibromyalgia: a population-based twin cohort study. BMC Musculoskelet Disord 2016; 17(1): 29.

32.

Creed

. The risk factors for self-reported fibromyalgia with and without multiple somatic symptoms: the Lifelines cohort study. J Psychosom Res 2022; 155: 110745.

33.

Landmark

Sunde

Fors

Kennair

LEO

Sayadian

Backelin

Reme

. Associations between pain intensity, psychosocial factors, and pain-related disability in 4285 patients with chronic pain. Sci Rep 2024; 14(1): 13477.

34.

Scherbov

Spitzer

Steiber

. Thresholds for clinical practice that directly link handgrip strength to remaining years of life: estimates based on longitudinal observational data. BMJ Open 2022; 12(7): e058489.

35.

Michaelides

Zis

. Depression, anxiety and acute pain: links and management challenges. Postgrad Med 2019; 131(7): 438–444.

36.

Choy

EHS

. The role of sleep in pain and fibromyalgia. Nat Rev Rheumatol 2015; 11(9): 513–520.

37.

Choi

Nah

Jang

Moon

Han

. Association between relative handgrip strength and chronic lower back pain: a nationwide cross-sectional analysis of the Korea National Health and Nutrition Examination Survey. Int J Environ Res Public Health 2021; 18(20): 10770.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.71 MB