The study was approved by the Griffith University Human Research Ethics Committee (ID: 2022/666) and written informed consent was obtained from all individuals. This cross-sectional investigation was conducted at the National Centre for Neuroimmunology and Emerging Diseases (NCNED) on the Gold Coast, Queensland, Australia. Eligible participants were contacted using the NCNED research registry database. ME/CFS patients were considered eligible if they fulfilled the CCC and/or ICC definitions for diagnosis, had received a formal diagnosis of ME/CFS by a physician, and did not report a history of COVID-19 infection. Participants with long COVID reported symptoms persisting for at least 3 months following COVID-19 infection according to the WHO working case definition. HC reported no diagnosis of a chronic health condition or evidence of underlying illness and had no current or prior COVID-19 infection. Participants were aged between 18- and 65-years. Medical history was requested to identify comorbid manifestations or exclusionary diagnoses including mental illness, malignancies, autoimmune, neurological, or cardiovascular diseases. Female participants were excluded if they were pregnant and/or breastfeeding. Finally, 10 ME/CFS patients fulfilling the CCC and ICC criteria (Carruthers et al., 2011), eight long COVID as defined by the WHO clinical case definition (World Health Organization [WHO], 2021) and 10 age-matched HC subjects were included in this study (see Table 1 for demographic information).

Symptom presentation was collected using the NCNED Research Registry questionnaire developed by NCNED with the Centres for Disease Control and Prevention (CDC) Symptom Inventory Questionnaire distributed online through LimeSurvey. The presence and severity of each symptom was assessed on a five-point scale: (1) very mild; (2) mild; (3) moderate; (4) severe; and (5) very severe. Validated patient-reported outcome measures were used to determine participant quality of life (QoL) and functional capacity. The 36-item short form health survey (SF-36) (Alonso et al., 1995) has been frequently employed in previous observational studies to assess QoL among people with ME/CFS (Eaton-Fitch et al., 2020), as well as, more recently, among people with the long COVID condition (O’Kelly et al., 2022). Eight QoL domains were assessed including physical functioning, role limitations due to physical health problems, bodily pain, general health perceptions, vitality, social functioning, role limitations due to personal or emotional health, and emotional wellbeing/mental health. Survey item scores were assigned a value between 0 and 100, before scores were averaged for each domain.

For subsequent correlation analysis, the severity measure of “pain” was extracted from SF36v2, while breathing scores were obtained via the NCNED Research Registry questionnaire. Symptom severity of 10 ME/CFS and eight long COVID patients have been provided as a Supplementary material.

Magnetic resonance imaging was performed on a 7 T whole-body MRI research scanner (Siemens Healthcare, Erlangen, Germany) with a 32-channel head coil (Nova Medical Wilmington, Wilmington, NC, USA). We acquired T1-weighted data using a Magnetization prepared 2 rapid acquisition gradient echo sequence (MP2RAGE) as in Thapaliya et al. (2019). In brief, MP2RAGE data were acquired sagittally using the following parameters: repetition time (TR) = 4,300 ms, echo time (TE) = 2.45 ms, first inversion time (TI1) = 840 ms, TI2 = 2,370 ms, first flip angle (FA1) = 5°, FA2 = 6° and resolution = 0.75 mm³ with matrix size = 256 × 300 × 320.

MP2RAGE data were processed similarly to our previous publications (Thapaliya et al., 2022a,b). In brief, MP2RAGE images were anatomically segmented using FreeSurfer version 7.1.1 (Fischl, 2012) ¹using the default FreeSurfer command “recon-all” on a Macintosh computer (Operating system: Catalina, RAM = 36GB, and core: 8). The “recon-all” processing includes motion correction, non-linear spatial normalization to Talairach space, intensity normalization, removal of non-brain tissue, cortical percolation, sub-cortical segmentation, gray and white matter boundary tessellation, automated topology correction, and surface deformation. Detailed information about the pipeline can be found at.²

Brainstem subregions were segmented using the FreeSurfer 7.1.1 brainstem module (Iglesias et al., 2015) as shown in Figure 1. Using this module, the brainstem was segmented into the midbrain, pons, superior cerebellar peduncle (SCP), and medulla oblongata. Brainstem subregions for all participants were visually checked for distortion-free segmentation.

Multivariate general linear model (GLM) statistical analysis was performed to test brainstem subregions and whole brainstem volume differences between ME/CFS, long COVID patients, and HC using SPSS version 28. After confirmation of homogeneity using Levene’s test, the multivariate GLM was used to test for three group differences. Correction for multiple group comparisons was implemented using the Bonferroni method. Then Spearman correlations were performed between brainstem subregion and whole brainstem volumes and clinical severity measures for ME/CFS and long COVID patients. The normality condition for data was checked using the Shapiro-Wilk method available in SPSS before the correlation. Age and sex were included as covariates for group comparisons and correlation analysis.

The brainstem subregion volumes were larger in ME/CFS patients compared with HC (see Table 2). After adjusting for multiple comparisons, volumes remained significantly larger in the pons (p = 0.004) and whole brainstem (p = 0.01) (see Figure 2 and Table 2).

In long COVID patients, after adjusting for multiple comparisons, we observed significantly larger volumes in the pons (p = 0.003), SCP (p = 0.009), and whole brainstem (p = 0.005) (see Figure 2 and Table 3). The medulla (p = 0.042) and midbrain (p = 0.026) volumes were not significantly larger compared with HC (see Table 3) after adjusting for multiple comparisons.

Although brainstem subregion volumes were smaller in ME/CFS patients compared with long COVID (see Table 4), these differences were not statistically significant (p < 0.05).

We demonstrated that subregion and whole brainstem volumes in ME/CFS and long COVID patients are significantly associated with clinical measures of “pain,” and “breathing difficulty” (see Figure 3 and Table 5). We observed a significantly strong positive relationship between “pain” and volume of pons (r = 0.83, p = 0.011) and whole brainstem (r = 0.85, p = 0.008) (see Table 5). There was also a strong negative relationship between “breathing difficulty” and midbrain (r = −0.78, p = 0.023) and whole brainstem (r = −0.78, p = 0.022) volumes in ME/CFS patients (see Figure 3). Furthermore, we found a very strong negative relationship between “breathing difficulty” and midbrain volume (r = −0.91, p = 0.03) in long COVID patients (see Figure 3 and Table 5).

Our study found significantly larger volumes for whole brainstem, pons, and SCP in ME/CFS and long COVID patients. The brainstem contains multiple small and dispersed neuron structures in the midbrain, pons, and medulla (Naidich et al., 2009) which together they constitute the reticular activation system (RAS). RAS nuclei connect with each other and to the body and subcortical and cortical structures (Guyton and Hall, 2011). RAS neurons influence cortical function via two different pathways. Firstly, RAS neuron projections deliver neurotransmitters directly or indirectly (e.g., via hypothalamus, basal forebrain) to the cortex (Saper and Fuller, 2017), and secondly RAS neurons generate oscillatory electrical signals that facilitate the coherence of cortical oscillations necessary for attention, sensory perception, problem solving, and memory (Garcia-Rill et al., 2013). Excitatory midbrain nuclei and inhibitory medulla nuclei constitute a circuit that controls both cortical arousal levels (cognition, wake/sleep, pain, respiration) and gait selection (e.g., walking or running) in response to inputs from multiple brain centers (Stornetta, 2008; Nicholls and Paton, 2009). Therefore, structural changes in the brainstem of ME/CFS and long COVID patients could result in severe and varied deficits in brain function.

We observed a larger volume for the whole brainstem and pons in ME/CFS patients compared with HC. Previous studies in ME/CFS patients have reported lower mean diffusivity in the pons (Thapaliya et al., 2021), and higher T1/T2 signal intensity in the medial lemniscus and cortical spinal tract (Thapaliya et al., 2020) that is sensitive to the level of myelination or iron. A functional MRI study reported impaired connectivity within the brainstem and to the hippocampus and thalamus of ME/CFS patients (Barnden et al., 2019). In ME/CFS patients, decreased myelin-sensitive T1-weighted spin echo signals were detected in the brainstem (Barnden et al., 2018) and the brainstem perfusion ratios were reduced (Costa et al., 1995). The brainstem contains the nuclei of the reticular activation system which control arousal, the sleep/wake cycle, gait, and memory via cortical connections and cardio-respiratory function (Costa et al., 1995; Garcia-Rill et al., 2013, 2016). Therefore, brainstem dysfunction is consistent with the symptoms experienced by ME/CFS patients including cognitive dysfunction, sleep disturbance, orthostatic intolerance, and dyspnea.

We also found larger volumes of the whole brainstem, pons, and SCP in long COVID patients compared with HC. Such volume increases may reflect edema of inflammatory responses, neurodegeneration, and/or viral invasion (Yong, 2021b). Autopsy studies of the brain have detected SARS-CoV-2 RNA and proteins in the brainstem of COVID-19 patients (Deigendesch et al., 2020; Matschke et al., 2020). Higher concentrations of SARS-CoV-2 are consistent with the high expression of Angiotensin-converting enzyme 2 which is the receptor SARS-CoV-2 uses to infect host cells in the brainstem (Letko et al., 2020; Zhou et al., 2020). Other autopsy studies showed inflammation, neuronal cell loss, and axonal degeneration in the brainstem of COVID-19 patients (Matschke et al., 2020; von Weyhern et al., 2020). Activated microglia and astrocytes, leukocyte infiltration, and micro-thrombosis have also been reported in the brainstem of COVID-19 patients (Deigendesch et al., 2020; Schurink et al., 2020; Meinhardt et al., 2021; Mukerji and Solomon, 2021). A microscopy study showed more tissue damage in the pons in COVID-19 patients than in controls (Bulfamante et al., 2021), and MRI also showed severe damage to the brainstem in two COVID-19 patients (Manganelli et al., 2020). Abnormal diffusion (lower fractional anisotropy) was reported in the SCP for multiple sclerosis patients with cerebellar symptoms and this correlated with cognitive performance (Nicoletti et al., 2017). The SCP has large sensory and motor nerve tracts that connect the cortex and pons and facilitate refined motor movements, learning of new motor skills, and balance (Khonsary, 2022). However, the function of this region needs to be investigated in different diseases. Damage to the brainstem, in particular the respiratory neurons of the dorsal medulla, could cause respiratory failure which is a key symptom of COVID-19 patients (Boutou et al., 2021a,b; Huang et al., 2021). Brainstem dysfunction has been demonstrated in chronic migraine headache (Aurora and Brin, 2017; Chong et al., 2017) which also occurs in long COVID (Membrilla et al., 2021). Therefore, structural changes in the brainstem are associated with the heterogeneous changes in brain function that correspond to the key symptoms of long COVID.

We did not find significant differences in the brainstem volumes of ME/CFS and long COVID patients which is consistent with the overlapping presentation of both cohorts (Sukocheva et al., 2021; Marshall-Gradisnik and Eaton-Fitch, 2022). Cardiovascular and respiratory symptoms of ME/CFS and long COVID are controlled by neuronal circuits between the hypothalamus and the brainstem (Benarroch, 2018). The symptom overlap between ME/CFS and long COVID patients is consistent with by our current findings of similar abnormalities in the brainstem. Further, a recent investigation demonstrated the biological overlap of ME/CFS and long COVID through transient receptor potential melastatin 3 (TRPM3) ion channel dysfunction (Sasso et al., 2022). TRPM3 ion channel dysfunction in the pathology of both ME/CFS and long COVID suggests further research is required to determine whether the illnesses are separate. TRPM3 channels are widely expressed through multiple cell and tissue types and are highly expressed in the brainstem, thus may account for a common pathology in ME/CFS and long COVID (Held and Tóth, 2021; Ragozzino et al., 2021).

We detected significant correlations between clinical measures (pain and breathing difficulty) and volumes of the whole brainstem and its subregions in ME/CFS and long COVID patients. Pain is regarded as one of the major symptoms of ME/CFS (Bourke et al., 2014). Our study shows a significantly strong positive correlation between “pain” and pons and whole brainstem volumes in ME/CFS patients (see Figure 3 and Table 5) indicating that larger brainstem volumes are associated with higher pain severity. The brainstem regions have several nuclei that receive ascending and descending signal pathways that inhibit or facilitate pain by upward or downward regulation of neurotransmission (Mills et al., 2021). Several brainstem nuclei including periaqueductal gray in the midbrain, dorsal and median raphe nuclei, parabrachial nucleus, and locus coeruleus in the pons region are involved in pain processing (Napadow et al., 2019). Functional connectivity differences were observed between brainstem nuclei in fibromyalgia patients (Ioachim et al., 2022). Recently, a study showed that the hippocampal subfield volumes were associated with pain levels in ME/CFS patients (Thapaliya et al., 2022b).

Breathing difficulty is another common symptom experienced by ME/CFS and long COVID patients (Ravindran et al., 2013), (Mancini et al., 2021). It has been reported that 30–50% of COVID-19 patients experience breathing difficulty (Mandal et al., 2021; Shah et al., 2021). We showed that smaller midbrain and whole brainstem volumes were associated with more severe “breathing difficulty” in both ME/CFS and long COVID patients (see Figure 3 and Table 5). Breathing difficulties in ME/CFS and long COVID are associated with brainstem volume changes that may reflect changes to the respiratory and cardiovascular neuronal circuits in the brainstem (Benarroch, 2018). The brainstem has a ventral respiratory column that controls rhythmic breathing (Smith et al., 1991; Moreira et al., 2011), a pontine respiratory group that controls the transition between expiration and inspiration (Stornetta, 2008), and the caudal ventrolateral medulla that controls inspiration (Nicholls and Paton, 2009). Therefore, brainstem dysfunction may contribute to the respiratory-related symptoms in ME/CFS and long COVID.

This study does have some limitations. This is a pilot study with a relatively small sample size that will affect the power of the study to detect brainstem volume differences and their association with clinical measures. Another limitation is that pain and breathing scores were obtained using self-reported questionnaires, which by their subjective nature may limit the interpretation of our findings. This study was a cross-sectional study; therefore, further investigations with a larger cohort and longitudinal studies are recommended to test progressive changes in the brainstem volume in ME/CFS and long COVID patients.