What this is
- This randomized controlled trial examined the effects of () on () in long COVID patients with fatigue.
- 42 participants with persistent fatigue for at least 12 weeks post-COVID were randomly assigned to either or sham treatment.
- Results indicated significant short-term improvements in metrics for the group, suggesting potential benefits for managing long COVID symptoms.
Essence
- significantly improved in long COVID patients with fatigue, indicating a potential therapeutic option for symptom management.
Key takeaways
- Osteopathic treatment increased the root mean square of successive differences (RMSSD) in , indicating enhanced parasympathetic activity post-treatment.
- The intervention group showed a decrease in the stress index and improvements in mean RR intervals, suggesting better autonomic regulation compared to the sham group.
- While the treatment effects were statistically significant, they were temporary, returning to baseline levels within 48 hours, indicating the need for further research on treatment frequency and long-term benefits.
Caveats
- The study's small sample size limits the generalizability of the findings, and the short follow-up period restricts conclusions about long-term efficacy.
- Self-reported fatigue improvements lacked established clinical significance, making it difficult to assess the true impact of the treatment on patient well-being.
Definitions
- Heart Rate Variability (HRV): A measure of autonomic nervous system function, reflecting the variation in time intervals between heartbeats.
- Osteopathic Manipulative Treatment (OMT): A hands-on therapy aimed at improving physiological function by addressing somatic dysfunctions through manual techniques.
Simplified
1. Introduction
Since the onset of the COVID-19 pandemic, an estimated 47% of the German population has been infected with SARS-CoV-2, with a case fatality rate of 0.47% [1]. While most patients recover from the acute phase, a substantial proportion continues to experience persistent symptoms, a condition now recognized as long COVID or post-COVID-19 syndrome. A large-scale prospective study reported that 13.3% of participants had symptoms lasting ≥28 days, 4.5% ≥8 weeks, and 2.3% ≥12 weeks after the initial infection [2]. Commonly reported symptoms include fatigue, dyspnea, musculoskeletal pain, and insomnia [3]. However, standard clinical investigations often fail to identify specific pathophysiological correlates [4]. Fatigue, in particular, is a leading complaint among long COVID patients and shares clinical features with post-infectious fatigue syndromes triggered by pathogens such as Epstein-Barr virus, influenza, enteroviruses, and intracellular bacteria [5,6]. Epidemiological data highlight the relevance of this issue: in a representative German population sample, the prevalence of severe fatigue symptoms was 14.5%, with higher rates among women and individuals with chronic illness or low socioeconomic status [7]. These findings underscore the urgent need for effective, low-risk treatment options for fatigue-related conditions. A growing body of evidence suggests that autonomic nervous system (ANS) dysfunction may contribute to the pathogenesis of long COVID symptoms. For instance, reduced parasympathetic activity and sinus tachycardia have been observed in 24 h heart rate variability (HRV) assessments of long COVID patients [8,9]. HRV is a well-established, non-invasive measure of autonomic balance and has emerged as a promising biomarker for dysautonomia and stress-related disorders. Osteopathic manipulative treatment (OMT) is a core component of osteopathic medicine, a holistic and patient-centered form of manual therapy that was developed in the late 19th century by the American physician Andrew Taylor Still. This approach is based on the idea that a person's structure and function are connected and that the body has self-regulatory mechanisms for healing. OMT encompasses various hands-on techniques that aim to improve physiological function by addressing somatic dysfunctions in the musculoskeletal, visceral, and cranial systems. A central tenet of osteopathy is enhancing autonomic regulation by manipulating fascia, joints, and soft tissues. OMT has been shown to have the ability to modulate the ANS and improve HRV parameters [10,11]. Recent studies suggest that OMT can increase parasympathetic activity—reflected in improvements in high-frequency (HF) power, RMSSD, and SDNN—and reduce sympathetic dominance, as indicated by decreased LF/HF ratios [12]. Given these physiological effects, OMT may offer a supportive therapeutic option for individuals with long COVID-related fatigue. However, empirical evidence for its efficacy in this context remains limited. The primary objective of this randomized, controlled, proof-of-concept trial was to evaluate whether OMT could induce measurable changes in HRV among long-haul COVID-19 patients with fatigue symptoms, using a sham-controlled design. In addition to HRV, the secondary objectives included assessing self-reported fatigue levels and subjective well-being before and after the intervention. We hypothesized that participants receiving OMT would demonstrate significantly greater improvements in HRV parameters, particularly increased parasympathetic activity, such as HF, RMSSD, and SDNN, but also decreased sympathetic dominance (LF/HF ratio). Furthermore, we expected OMT to be associated with greater reductions in fatigue severity and improved self-perceived health status.
2. Methods
2.1. Study Design
This single-center, randomized, controlled trial (RCT) was conducted in accordance with the ICH-GCP guidelines and the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the Department of Human Medicine at Philipps University Marburg (Ref. No. 23-08 BO). The study was prospectively registered in the German Register of Clinical Trials (DRKS-ID: DRKS00030829) and is listed on the WHO International Clinical Trials Registry Platform (ICTRP). Participant recruitment and data collection, including all HRV measurement time points, were conducted under supervision between June and November 2023 at the Schoen Klinik Berchtesgadener Land, Schoenau am Koenigssee, Germany. All participants provided written informed consent. Randomization was performed by a staff member not involved in the intervention or data analysis. The randomization process was executed through the utilization of sealed envelopes, which were arranged into three blocks of ten and two blocks of six numbers, without the predetermined block characteristics. An independent and blinded third party was designated to reveal the group assignment upon request and link it to the respective study identification code. Outcome assessors and data analysts were blinded to group allocation.
2.2. Study Procedures
The study comprised three time points: baseline (T1), immediately after the intervention (T2), and 48 h follow-up (T3; Figure 1). Prior to inclusion, each participant underwent a standardized medical evaluation to confirm a diagnosis of long COVID-19 syndrome according to Koczulla et al. (2021) [4]. Participants completed a medical history questionnaire, the Fatigue Assessment Scale (FAS), and a subjective dyspnea questionnaire assessing exercise-induced breathlessness on a scale from 0 (no impairment) to 10 (unbearable). To assess the integrity of blinding, participants completed a de-blinding questionnaire at follow-up, which asked whether they believed they had received an actual treatment or placebo, based on prior protocols [13,14].
2.3. Participants
A total of 42 participants were included (Figure 1). No subjects were excluded during enrollment. Inclusion criteria were: age ≥ 18 years; documented SARS-CoV-2 infection with persistent symptoms for ≥4 weeks; and an FAS score ≥22. Participants were excluded if they had previously received osteopathic treatment for fatigue or long COVID symptoms. Further exclusion criteria included current alcohol or drug abuse (due to its influence on HRV) and first-trimester pregnancy due to limited data on intervention safety in this group [15]. The study included subjects who were over 18 years old at the time of data collection. The patients had previously recovered from an acute SARS-CoV-2 infection and were in the post-acute phase at least 4 weeks afterward, as defined in the NICE guidelines [3]. Since low HRV values are typically observed in patients with fatigue syndrome [16], patients experiencing fatigue were included to ensure baseline value homogeneity. Fatigue is the most common symptom in long COVID, accounting for 58% of patients [17]. The FAS was used to confirm the presence of fatigue, and patients with a cutoff score of 22 or higher were included in the study. All patients underwent a conventional medical differential diagnosis at the start of their rehabilitation to confirm the LPCS diagnosis [4,18]. To best ensure blinding of the control group receiving the sham treatment, subjects who had not previously received osteopathic treatment for this condition were included. Subjects with a history of alcohol or drug abuse were excluded. Such abuse can reduce HRV in the short and long term, which could influence the analysis [15]. Pregnant women in the first trimester were also excluded.
2.4. Osteopathic Treatment Intervention
2.4.1. Osteopathic Treatment Group
Participants in the intervention group received a single 45 min session of osteopathic treatment. A standardized assessment was performed in standing, sitting, and supine positions using the Zink test to identify somatic dysfunctions (Figure 1). Diagnoses were based on tissue texture abnormalities, asymmetry, range of motion restrictions, and tenderness. Treatment was individualized according to osteopathic findings and applied using techniques deemed appropriate by the therapist. Interventions were performed by two certified osteopaths with extensive clinical experience.
2.4.2. Sham-Treatment of Control Group
Control group participants received a 45 min sham treatment designed to mimic the structure of the OMT session (Figure 1). Practitioners placed their hands on various body regions without applying therapeutic manipulations and silently counted backward to simulate treatment duration and engagement [12,19]. No verbal feedback or clinical findings were shared during or after the sessions. This approach has been validated in previous studies examining autonomic nervous system outcomes [13,14].
2.5. Outcome Measures
Heart rate variability (HRV) was recorded for 10 min using the Faros 180 ECG device (Bittium GmbH, Oulu, Finland) at each time point. Participants were instructed to remain awake, breathe calmly (approx. 9–20 breaths/min), and avoid physical activity before measurement (Figure 1). Data were analyzed using Kubios HRV software 4.0.2 (Kubios Oy, Kuopio, Finland), which supports multiple ECG input formats and computes both linear and non-linear HRV parameters [20]. Measured parameters included time-domain metrics (RMSSD, mean RR, mean HR, SDNN, pNN50, NN50), frequency-domain metrics (LF, HF, LF/HF ratio), and non-linear dynamics (Poincaré plot). The Baevsky Stress Index, representing sympathetic nervous system activity, was also calculated based on RR interval variability [21].
2.6. Data Analysis
Sample size was estimated using G*Power software 3.1.9.7 [22]. Based on a similar study [11], we used the smallest reported effect size (ƞp2 = 0.117), an alpha error level of 0.05, and a power of 0.8 to be on the safe side. Using these parameters for the F-test, we obtained a total sample size of n = 42. This corresponds to 21 individuals in each group. Normality was assessed using the Shapiro–Wilk test; homogeneity of variances via Levene's test. For repeated measures, a robust mixed ANOVA was conducted to evaluate interaction effects (group × time). In addition, a multistep, estimation-oriented approach was applied to assess intragroup and intergroup effects across time points. For each variable, descriptive statistics and standardized effect sizes (Hedges' g) were calculated, including 95% confidence intervals (CI) via bias-corrected bootstrapping (1000 resamples); for group differences, small, medium, and large effect sizes correspond to 0.1, 0.4, and 0.8, respectively [23]. This enabled a reliable interpretation of temporal trends and effect magnitudes, accounting for deviations from normality and unequal variances. To validate findings and estimate individual variability over time, a simulation-based bootstrap procedure (B = 1000) was used to generate empirical confidence envelopes for mean HRV trajectories. This approach follows APA recommendations [24,25] by prioritizing effect size estimation over sole reliance on p-values [26]. Baseline comparability between the osteopathy and sham groups was assessed using appropriate tests. These tests confirmed no significant baseline differences between groups across demographic and clinical parameters (see Table 1).
3. Results
3.1. Symptoms Associated with Long COVID and Fatigue Symptoms
A total of 42 participants were included in the study, with ages ranging from 23 to 75 years (Table 1). The mean age was 50.1 ± 11.7 years in the sham group and 51.0 ± 12.5 years in the osteopathy group. Overall, 26% (n = 11) of participants were male and 74% (n = 31) female. A Sankey diagram (Figure 2) illustrates the distribution and overlap of treatment group (OMT vs. sham), sex, vaccination status, and prior COVID-19-related hospitalization. The width of each stream corresponds to the number of individuals in the respective category combination. The mean time from initial SARS-CoV-2 infection to the study intervention was 15.1 ± 7.5 months. In the sham group, this interval was 12.5 ± 1.2 months, compared to 17.2 ± 1.9 months in the OMT group. Most participants (n = 36) had been vaccinated prior to their acute infection, with 19 in the sham group and 17 in the OMT group. Detailed information on acute symptom profiles is provided in Table 1.
3.2. Stress and Heart Rate Changes in OMT Treatment
Compared to the control group receiving sham treatment (placebo control), the osteopathy group exhibited consistent and temporally structured changes across key heart rate variability (HRV) indices. RMSSD significantly increased from t1 (mean: 17.8 ms ± 11.2, n = 15) to t2 (mean: 21.6 ms ± 9.3; g = 0.82, 95% CI [0.19, 1.40]) and declined by t3 (mean: 17.4 ms ± 7.8; g = −0.55, 95% CI [−0.96, 0.18]; Figure 3A). A similar pattern was observed in the stress index, which initially decreased from t1 (mean: 16.2 ± 6.2, n = 18) to t2 (mean: 12.8 ± 5.0; g = −1.02, 95% CI [−1.33, −0.59]) and subsequently increased again at t3 (mean: 18.0 ± 6.8; g = 0.94, 95% CI [0.56, 1.34]; Figure 3B). The SDNN increased significantly from t1 (mean: 19.6 ms ± 8.6, n = 17) to t2 (mean: 26.0 ms ± 12.5; g = 0.80, 95% CI [0.38, 1.18]) and declined at t3 (mean: 21.4 ms ± 10.3; g = −0.51, 95% CI [−0.92, 0.01]; Figure 3C).
Mean RR intervals increased markedly between t1 (mean: 869 ms ± 104, n = 19) and t2 (mean: 934 ms ± 102; g = 1.01, 95% CI [0.43, 1.61]) and returned to lower values by t3 (mean: 857 ms ± 104; g = −1.04, 95% CI [−1.56, −0.52]; Figure 3D). The inverse pattern emerged for mean HR, which decreased from t1 (mean: 67.3 bpm ± 7.8, n = 18) to t2 (mean: 62.3 bpm ± 5.9; g = −1.10, 95% CI [−1.62, −0.48]) and increased again at t3 (mean: 68.8 bpm ± 7.1; g = 1.03, 95% CI [0.53,1.57]; Figure 3E).
In the frequency domain, low-frequency power (LF) increased significantly from t1 (mean: 206 ms2 ± 153, n = 17) to t2 (mean: 329 ms2 ± 277; g = 0.53, 95% CI [0.07, 0.86]) and remained stable by t3 (mean: 247 ms2 ± 217; Figure 3F). High-frequency power (HF), as an index of parasympathetic tone, rose from t1 (mean: 146 ms2 ± 207, n = 14) to t2 (mean: 213 ms2 ± 202; g = 0.48, 95% CI [−0.14, 0.99]) and dropped at t3 (mean: 119 ms2 ± 91; g = −0.59, 95% CI [−0.95,−0.27]; Figure 3G).
The LF/HF ratio remained statistically unchanged across all time points within the osteopathy group. In contrast, the control group exhibited a moderate decrease in this ratio between t2 (mean: 2.05 ± 0.53) and t3 (mean: 1.09 ± 0.28; g = −0.53, 95% CI [−0.83, −0.04]), as well as a significant decline in mean RR intervals (g = −0.79, 95% CI [−1.43, −0.03]) in the same interval (Figure 3H). Total power (TP) rose from t1 (mean: 487 ms2 ± 457, n = 18) to t2 (mean: 822 ms2 ± 773; g = 0.62, 95% CI [0.35, 0.89]) but did not increase further at t3 (mean: 561 ms2 ± 534; g = −0.36, 95% CI [−0.74, 0.12]; Figure 3I). All other metrics in the sham group remained largely stable.
Five outliers were identified in the dataset, four of them in the osteopathy group, and two in the sham group (Figure 4; Table 2). The osteopathic group showed an increase in RMSSD values at t2, followed by a significant decrease from t2 to t3. In the sham group, on the other hand, the values hardly changed. Due to the small sample size, no effect sizes were calculated.
At the end of the study, all participants completed a de-blinding questionnaire. In the sham group, 13 participants correctly assumed that they were in this group, whereas eight thought that they had received osteopathic treatment. This results in a percentage ratio of 61.9% to 38.1%. In the de-blinding questionnaire, respondents were asked how strongly they believed that treatment had taken place and were rated on an NRS of 0–10. The participants who had seen themselves in the osteopathy group gave a mean value of 6.7 ± 2.6, and those who had seen themselves in the placebo group 5.8 ± 2.9.
4. Discussion
Persistent symptoms beyond the acute phase of SARS-CoV-2 infection—lasting weeks to months—are referred to as long COVID-19 syndrome [3,4,27]. In our study cohort, participants presented with fatigue-related symptoms and autonomic dysfunction, paralleling findings observed in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). In ME/CFS, significantly reduced HRV parameters—such as mean RR, SDNN, RMSSD, and pNN50—have been reported, reflecting reduced vagal tone and increased sympathetic activity [16]. Similarly, frequency domain analyses in ME/CFS revealed decreased LF and HF power and an elevated LF/HF ratio, indicating autonomic imbalance. Osteopathic manual treatment (OMT) has been suggested as a potentially beneficial intervention for patients with fatigue syndromes, including ME/CFS [28] and, more recently, for those with long COVID [29]. In the present study, we observed transient yet consistent changes in HRV following OMT: RMSSD and SDNN significantly increased, and the stress index declined shortly after treatment (t2), indicating acute improvements in parasympathetic activity. These changes were reversed by the follow-up time point (t3), consistent with prior observations of short-term autonomic modulation after therapeutic interventions [30]. This temporal pattern aligns with previous studies: in COVID-19 survivors, RMSSD was observed to increase during early recovery, and similar post-treatment changes have been reported in ME/CFS [31]. However, HRV findings in post-COVID-19 cohorts have been inconsistent across studies [32], with reports of both increased and decreased HRV values [8,33], likely due to differences in timing, symptom severity, and measurement protocols. The observed acute HRV improvement after OMT suggests that the intervention temporarily modulates autonomic tone—potentially via vagal activation—without inducing long-lasting effects after a single session. This is supported by changes in time and frequency domain parameters, including RMSSD, SDNN, mean RR, HF power, and total power, which peaked post-treatment and returned to baseline within 48 h. Notably, stress-related markers (e.g., mean HR, stress index, LF/HF ratio) also showed a reversal between t2 and t3, indicating a return to sympathetic predominance. Importantly, the sham group showed no comparable changes in HRV parameters, supporting the hypothesis that the effects observed in the OMT group were not due to placebo or expectation alone. This is in line with earlier findings indicating the potential of manual therapies to induce acute autonomic modulation [10,34]. Nevertheless, nonspecific treatment effects such as patient expectations, therapeutic touch, and the general experience of attentive care may have contributed to the observed responses. The impact of placebo effects in manual therapy research has been thoroughly documented, particularly in conditions such as fatigue, where subjective perceptions play a pivotal role in outcome reporting [35,36,37]. Despite the meticulous design of our sham intervention to control for these factors, the possibility of residual expectancy effects cannot be entirely discounted, particularly in light of the open nature of the patient-therapist interaction. The concept of therapeutic touch has been demonstrated to impact autonomic responses, potentially through mechanisms involving affective touch, interoception, and oxytocin release [38]. These aspects are not exclusive to OMT and could potentially partially explain short-term changes in HRV. Five participants in our study exhibited unusually high RMSSD values, deviating from the expected profile of fatigued patients, where RMSSD is typically reduced [16]. Interestingly, 75% of these outliers had been hospitalized during the acute phase of COVID-19, consistent with reports linking post-hospitalization to elevated HRV [32]. However, the possibility that psychiatric comorbidities, such as anxiety disorders—which can paradoxically increase HRV [39]—contributed to these findings must be considered. The use of standardized psychometric assessments in future studies is, therefore, recommended. While the presence of large standard deviations in HRV parameters, such as RMSSD, LF, and HF, underscores high interindividual variability [32,40], the consistent group-level patterns and medium-to-large effect sizes support the reliability of the observed responses to OMT. Factors such as breathing rate, stress perception, and baseline autonomic tone may modulate individual responses [41].
These findings emphasize the importance of person-centered analytic strategies—such as responder classification or latent class modeling—to identify subgroups that benefit most from OMT. Such approaches could enhance therapeutic precision in future studies [42]. It is crucial to consider the study as a pilot one. While the effects of one OMT session appear promising, the results' limited applicability is shown by the small sample size and brief follow-up period. The observed changes are too temporary to draw definitive therapeutic conclusions yet; OMT might help in the long-term effects of post-COVID syndrome when combined with other treatments. Research is needed on OMT's role in this context. Previous studies on similar approaches for post-viral syndromes have shown mixed results, so more research is necessary to clarify OMT's role in broader therapeutic settings [43]. However, the study is not without limitations. The most critical are the small sample size, the absence of long-term follow-up, and the exclusive focus on HRV as an outcome measure. Moreover, the observed return of HRV parameters to baseline within 2 days limits conclusions about sustained therapeutic effects. Self-reported fatigue severity was reduced, but the minimal clinically important difference (MCID) for the Fatigue Assessment Scale (FAS) has not been established in long-COVID populations. The clinical relevance of these changes is unclear, so they should be interpreted with caution. Future studies should determine the MCID for FAS in this context to better interpret therapeutic impact. Future research should involve larger cohorts, repeated treatment sessions, and multimodal outcomes (e.g., fatigue scales, quality of life measures) over extended periods. This study is the first to systematically evaluate the effects of OMT in long COVID-19 patients using robust HRV analysis in a controlled trial setting. The results indicate that OMT can transiently enhance parasympathetic tone and modulate autonomic function in affected individuals. Although the effects appear short-lived, they suggest a therapeutic window for intervention. Nonetheless, it must be noted that these effects were short-lived and should be interpreted as preliminary rather than conclusive evidence for therapeutic efficacy. In the broader context, the rhythmic nature of HRV in relation to biological cycles—such as sleep, activity, and hormonal rhythms—aligns conceptually with the rhythmic principles of osteopathy. Techniques that aim to synchronize craniosacral and tissue rhythms could offer new avenues for monitoring and optimizing treatment outcomes through long-term HRV assessment.
In conclusion, while this study demonstrates that OMT can induce short-term autonomic improvements in patients with long COVID-19, the return to baseline by t3 highlights the need for repeated treatments and longitudinal research. Future studies should evaluate the durability of these effects and assess whether sustained OMT interventions can contribute to long-term recovery in individuals with post-viral autonomic dysfunction.
Long Covid