Chronic Fatigue Syndrome and Cardiovascular Disease

JACC State-of-the-Art Review

Benjamin H. Natelson, MD; Danielle L. Brunjes, PHD; Donna Mancini, MD


J Am Coll Cardiol. 2021;78(10):1056-1067. 

In This Article

Connection to Cardiovascular Disease

Cardiac Dysfunction

Reduced Stroke Volume. Eleven ME/CFS patients, selected because they had runs of flat or inverted T waves on 24-hour Holter,[12] underwent stress multigated acquisition and were found to have abnormalities in ejection fraction, wall motion, or cardiac size.[13] A later study done by Natelson's group using impedance cardiography found reduced stroke volume in ME/CFS patients on the severe end of the illness spectrum as determined by severity and frequency of symptoms when compared with data from sedentary, matched healthy control subjects.[14] Of particular interest was the fact that a significant inverse relation was found between cardiac output and self-related severity of PEM such that patients with progressively more burden with PEM had progressively reduced cardiac output. Because blood pressure was maintained in these patients, the conclusion was that a low output circulatory state existed for some patients.

Researchers in Miami, Florida, confirmed this finding of reduced stroke index in ME/CFS patients who were also on the severe end of the illness spectrum—again compared with sedentary healthy control subjects.[15] The researchers used the operational criteria for "severe ME/CFS" that Natelson's group had previously developed;[14,16] approximately one-half of the patients studied were in this severe group. Importantly, the Miami group found that this reduction was a function of reduced blood volume, defined as 8% decrease from normal total blood volume and not a problem with cardiac contractility. In their analysis, the observed group differences for cardiac volume were corrected by controlling for total blood volume deficits. The frequency with which subjects had reduced total blood volume was 63% in the severe ME/CFS group, 27% with nonsevere ME/CFS, and 0% in sedentary control subjects; thus, about one-half of the patients had low blood volume. The authors suggested that a blood volume deficit affected oxygen delivery and nutrient supply, thus impairing hemodynamic regulation, resulting in fatigue and other symptoms of ME/CFS. Another group extended these findings, reporting not only lower red blood cell volume in ME/CFS than in control subjects and significantly reduced stroke, end-systolic, and end-diastolic volumes, but also reduced end-diastolic wall mass in the ME/CFS group as determined using cardiac magnetic resonance.[17] A good correlation between wall mass and total volume and a significant negative relation between plasma volume and fatigue severity was further described by these investigators.

One possible explanation for the reduced ventricular wall mass in ME/CFS may be caused by overall heart size. Indeed, 61% of a group of Japanese ME/CFS patients were reported to fulfill an operational definition of small heart size—calculated cardiothoracic ratios ≤42% from A-P view of chest x-ray. Mitral valve prolapse was identified by echocardiography in 29% of those with small hearts compared with 0% in patients whose cardiothoracic ratio exceeded 42%.[18] Cardiological evaluation of these patients found abnormalities: first, about one-third of patients reported dyspnea on exertion or chest pain; one-fifth had ECG evidence of right axis deviation including vertical axis (electrical axis 90°); and they had significantly reduced stroke volume and cardiac indexes compared with healthy control subjects.[19] Some of these patients showed clinical improvement, at which time follow-up evaluation showed larger heart size and improved stroke volume/cardiac output; the obvious inference was a relation between cardiac dysfunction and illness severity.[20] Another group using magnetic resonance spectroscopy found a trend to reduced creatine phosphate/adenosine triphosphate compared with healthy control subjects; these authors dichotomized patients into a normal cardiac metabolic group and low creatine phosphate/adenosine triphosphate groups, with those in the impaired cardiac group showing significantly reduced both maximal and initial proton efflux rates.[21] And yet, another study noted reduced cardiac mass, stroke volume, ejection fraction, and end-diastolic diameter in ME/CFS.[22]

Importantly, most of the studies reviewed in the previous text did not use sedentary healthy people as control subjects for the cardiovascular changes reported, and deconditioning could explain some of these findings. This seems to be the case, at least in part, related to the question of reduced plasma volume and small heart size. Levine's group reported that 2 weeks of head-down bed rest produced similar cardiovascular changes: namely, reduction of stroke volume, plasma volume, and ventricular mass.[23] Not using such sedentary control subjects seems to explain the chronotropic incompetence reported several times over the years in ME/CFS patients performing a cardiopulmonary exercise test (CPET).[24,25] Cook et al[26] matched ME/CFS patients with healthy, sedentary control subjects based on actual maximal VO2 attained; with this careful matching for fitness, heart rates during CPET were the same for patients and control subjects;[26] this effect has recently been replicated in a larger study where 99 ME/CFS patients were similarly matched to 99 healthy control subjects—again showing no difference in heart rate over time (Dane Cook, personal communication, May 1, 2021). Because Vo2max values were the same for the 2 groups in both these studies, they are the first in which one can infer that deconditioning is not responsible for the fatigue and other symptoms seen in ME/CFS.

Although such data do not support the idea of deconditioning as a causal factor in the genesis of the symptom complex of ME/CFS, deconditioning probably can increase any patient's symptom burden. Inactivity-induced deconditioning can result in a reduction of total blood volume that affects stroke volume and leads to increased sympathetic and reduced parasympathetic activity to the sinus node. Cardiac atrophy can occur in as little as 2 weeks of marked inactivity.[23] In addition, lack of activity can produce muscle atrophy, with decrease in muscle strength and endurance, enzymatic changes, decrease in capillarization, changes in body composition, changes in plasma volume, and changes in autonomic function. Moreover, deconditioning can lead to even more inactivity, which further exacerbates deconditioning resulting in a downhill spiral. Therefore, deconditioning imposed by the rest inherent in having ME/CFS may explain some of the putative cardiac abnormalities previously reported.

Interestingly, a recent publication on healthy aging and cardiovascular function describes the hemodynamic response to exercise over 6 decades of life.[27] Functional and structural changes occur with aging, which lead to a decrease in resting cardiac output and, because resting heart rate is generally unchanged, to a lower stroke volume. In healthy older adults, peak VO2 was reduced because of lower stroke volume, peak heart rate, and peak exercise end-diastolic volume in the presence of similar ventricular filling pressures, suggesting reduced inotropic reserve or possibly reduced Frank-Sterling reserve. Therefore, the exercise response of patients with ME/CFS is not dissimilar from normal aging.

Oldham et al[28] described the exercise hemodynamic response in a group of 619 patients with unexplained dyspnea. They identified 49 patients with low peak VO2, with normal biventricular function but low peak cardiac output, and without exercise-induced pulmonary hypertension. These patients appeared limited by their inability to increase their filling pressures during exercise, and the authors hypothesized that there was a failure of their Frank Starling response possibly from inadequate venous vasoconstriction. A subgroup of 23 patients underwent fluid loading with normal saline followed by a second exercise test. The fluid loading resulted in increase in cardiac output and VO2 in 70% of these patients. Although these patients were not identified as having ME/CFS, the inference is that many may have had that diagnosis. Interestingly, these authors report that deconditioned subjects had higher pulmonary capillary wedge and RA pressures at matched workloads than trained counterparts, although their CO and SV were lower. Patients with deconditioning did not exhibit the preload failure that was observed in the patients who would be most similar to ME/CFS patients.

Reduced Blood Pressure

Blood pressure abnormalities have been variably described in this disorder. Although 1 study found no differences in blood pressure between patients and control subjects,[29] other studies described reduced blood pressure as assessed by 24-hour monitoring with reductions in both systolic and diastolic pressure[30] or reduction in systolic pressure alone.[31] This reduction in blood pressure was more marked during the nighttime than the daytime, and there was a modest relation between increased fatigue severity and the variability between daytime and nighttime SBP over 24 hours.[30] These researchers concluded that "low BP and regulation of BP particularly may contribute to the manifestation of the symptom of fatigue".[30] Although this group did not use sedentary healthy people as control subjects, they did use patients with fatigue secondary to primary biliary cirrhosis who differed from those with ME/CFS by showing only a modest, but significant reduction in systolic blood pressure over the 24-hour day. Another study noted that patients had stiffer arteries, which correlated both with systolic blood pressure and levels of C-reactive protein.[32] The autonomic dysfunction commonly described in these patients may also contribute to variations in blood pressure control. This will be discussed below.

Autonomic Dysfunction

There has been a lot of interest in the role of autonomic dysfunction in ME/CFS. Pain and fatigue experienced by patients are often correlated to symptoms of autonomic dysfunction. One research method of capturing autonomic activity[33,34] is by assessing heart rate variability (HRV) which captures the magnitude of respiratory-coupled sinus arrythmia—a variable that allows quantification of parasympathetic modulation of heart rate. Reductions in this variable have been shown to predict all-cause death and cardiovascular events.[35] The relation between sympathetic activity and HRV is not as clear-cut, although some inferences can be drawn using statistical approaches to capture slower changing bands of HRV.

Natelson et al have done a number of studies to evaluate HRV in CFS. In the first of these studies using paced breathing—a method that magnifies respiratory-coupled, parasympathetic activity—ME/CFS patients showed reductions in both sitting and standing positions compared with healthy control subjects.[36] In a subsequent study using the orthostatic challenge of head up tilt, ME/CFS patients had a lower mean RR interval compared with sedentary control subjects,[37] suggesting that autonomic dysfunction is present during times of gravitational stress. Several studies have shown evidence of decreased HRV during sleep in ME/CFS patients.[38,39] In the study by Togo and Natelson,[38] autonomic dysfunction during non-REM sleep was found to correlate to unrefreshing sleep—a relation that could exacerbate the common symptom of daily fatigue in ME/CFS patients. A meta-analysis[40] supported these findings, noting that ME/CFS patients had a higher ratio between low- and high-frequency power (thought to reflect sympathetic activity) than control subjects as well as lower high-frequency power. An important recent study showed a good correlation between measures of fatigue and a number of HRV variables.[41] This result suggests that HRV could be a useful objective outcome measure in future clinical trials.

Orthostatic Intolerance

Another symptom common in ME/CFS is the report of worsening symptoms (more cognitive problems, fatigue) and/or blurred vision and lightheadedness while upright—ie, self-reported orthostatic sensitivity. This symptom may reflect both the aforementioned reduction in blood volume as well as impairment in autonomic control. A recent paper reported that self-reported symptoms consistent with OI occurred in 86% of patients.[42] Patients with these symptoms have been shown to show reductions in cognitive function following orthostatic challenge.[43] Should sympathetic excitation exist (a subset of those with OI), other symptoms might include palpitations, chest pain, and tremulousness. Although approximately one-half of patients with these reported symptoms have no physiological abnormalities during orthostatic challenge, the abnormality most often reported is postural orthostatic tachycardia syndrome (POTS). POTS is identified as when a subject's supine, resting heart rate increases by 30 beats/min or if the subject's heart rate exceeds 120 beats/min. Several papers have noted POTS to be common in ME/CFS related to severity of fatigue, and are therefore important to be recognized.[42,44] In contrast, another paper reported found no difference in prevalence when compared with healthy control subjects.[45]

POTS has been described in a very heterogenous group of diseases besides ME/CFS, with multiple different pathophysiological subtypes that include neuropathic, hypovolemic, hyperadrenergic, and Ehlers Danlos Syndrome-hypermobility related phenotypes. Although POTS is common in ME/CFS, not every patient with POTS has ME/CFS. We view the existence of POTS and any other physiological manifestations of OI as a way to stratify ME/CFS patients; the inference we make is that those with physiological manifestations of OI may have a different underlying pathophysiological process producing their ME/CFS symptoms compared with other patients without OI.

Several studies have compared ME/CFS with POTS to those without this comorbidity. One study found that those with POTS were younger, reported less fatigue and sleepiness and had lower depression scores than the non-POTS patients.[46] Another study was similar but reported higher levels of fatigue, shorter illness duration, and lower systolic blood pressure during orthostatic challenge.[47] In another study, those with POTS had higher plasma renin activity in both supine and upright positions.[48] In contrast, another study in POTS[49] reported that despite the lower plasma volume found in these patients, a compensatory increase in plasma renin and aldosterone activity was not seen. Finally, one recent study found those with reports consistent with OI had lower blood volumes than those without reported symptoms.[50]

Concerning age, the rate of POTS in adolescents with ME/CFS approaches 100%,[51] but it is seen much less often in adults.[5] Of interest is a recent report of reduced cerebral blood flow during head up tilt in ME/CFS—even when there was no evidence of OI.[42]

Earlier reports noted a high rate of orthostatic hypotension occurring late in a 45-minute head-up tilt.[52] Two groups used the same protocol but were unable to confirm differences in blood pressure between patients and healthy control subjects during the same 45-minute head up tilt procedure.[53,54] This orthostatic hypotension is not seen often during standardized 10-minute orthostatic challenges except for in patients taking antihypertensives. In our experience, fainting is uncommon using the 10- minute challenge.

To simplify the assessment of patients for OI, we have adapted the 10-minute lean test originally used by the National Aeronautics and Space Administration to evaluate returning astronauts for this problem.[55] To do this, we collect supine blood pressure, heart rate, respiratory rate, and end-tidal CO2 once per minute for several minutes after giving the patient adequate time for these physiological markers to become baseline; then, we ask the patient to lean against a wall with feet together touching only shoulder blades to the wall and collect these same variables once per minute for 10 minutes.

Using this same 10-minute test, 1 recent paper has noted that reduced pulse pressure occurs during the latter one-half of the lean test in patients sick for <4 years; this is thought to be an indirect measure of reduced blood volume.[56] One of the reasons that some groups have not identified POTS as a common problem in ME/CFS may be because an alternate form of OI can exist. Specifically, several groups have reported the postural orthostatic syndrome of hypocapnia (POSH).[5,57] Our own study indicated that POSH occurs more often than POTS.[5] Ocon et al[58] have shown that POSH occurs first and is the driver for the development of POTS. These researchers also have shown that, while undergoing orthostatic challenge, the cognitive processing abilities of young adults deteriorates[58] but with no change in cerebral blood flow. An important negative is that anxiety was ruled out as a factor in producing either POTS[59] or POSH.[5]

The mechanism of OI continues to be an important target for research. The change in posture with orthostatic challenge leads to a shift in plasma volume with venous pooling caused by gravity leading to a transient decline in venous return. This results in a decrease in cardiac output and blood pressure with 2 consequences—unloading of the baroreceptors with subsequent sympathetic stimulation to increase heart rate, vasoconstriction, and venous return and/or activation of chemoreceptors that produce hyperpnea that functions possibly to pull blood into the chest. These compensatory mechanisms are impaired in POTS and POSH.

One possible cause of these dysregulated responses is systemic hypovolemia, which was reported to occur in ME/CFS many years ago by Streeten and Scullard.[60] Joyner and Masuki[61] concluded that the physiological manifestations of OI are identical to what is seen under conditions producing marked deconditioning. This conclusion must be tempered by the need for studies in which the physiological response to orthostatic challenge in patients is compared with that found in matched sedentary control subjects. Importantly, data do exist that a program of gentle physical conditioning can improve work capacity in ME/CFS as well as self-reported health.[62]

Another risk factor for OI—especially for POTS—is the coexistence of the hypermobile form of Ehlers-Danlos syndrome.[63] This is operationally defined by Beighton score testing as hyperflexibility in 5 sites of the body plus the report of arthralgia in at least 4 joints. Further assessment includes a positive family history of hypermobility with joint pain reports as well as a history of musculoskeletal problems (eg, long-term pain, dislocations) and signs of faulty connective tissue throughout the body (eg, hypermobile skin, hernias, prolapses). Heart and blood vessel conditions occurring with EDS include heart valve and vessel dysfunction, including mitral valve prolapse and aortic root dilation. Fatigue is acknowledged to be a common report of patients with Ehlers-Danlos syndrome,[64] and, accordingly, over 82% of hypermobile EDS patients fulfilled diagnostic criteria for ME/CFS.[65]

As a group, ME/CFS patients have increased blood norepinephrine levels compared with control subjects on orthostatic challenge.[66] When baroreceptor activation produces excessive sympathetic activation, a hypertensive response to orthostatic challenge can occur—that is, in a patient with no history of hypertension. This "hyperadrenergic" form of POTS may have other cardiac-related symptoms.[67] Gibbons et al[68] used the existence of small fiber neuropathy and raised sensory thresholds in the foot to define a group of POTS patients as being neuropathic. They note that those in the non-neuropathic group show evidence of increased sympathetic activity on Valsalva testing.

These studies on mechanisms suggesting subgroups of causes of OI raise questions for future research. One critical problem is the lack of a study evaluating all of the possible subgroups that includes a large series of patients with reports indicative of orthostatic testing. For example, the possibility that patients with normal biopsy for small fiber neuropathy are those who show an exaggerated blood pressure response to head up tilt—ie, the hyperadrenergic form. And, of course, an unanswered question is if the blood volume of patients in these groups is normal or reduced.

Postexertional Malaise

A key characteristic of ME/CFS is worsening across the broad range of symptoms constituting ME/CFS after relatively minor physical or even cognitive efforts. This PEM is common in ME/CFS and is thought to be a sine qua non for the illness.[3,69] Surprisingly, most data on PEM derive from patient self-report collected at a time unrelated to an actual bout of PEM. In 1 study, 11% of patients reported a delay of at least 24 hours, with most reporting symptoms lasting over a day.[70] Another study, using an epidemiological approach, noted that the majority of patients reported symptom worsening occurring within 2 days of exertion, and only 9% reported a delay of at least 5 days before this happened.[71] One group evaluated patients for PEM a week after they performed CPET; the patients retrospectively reported having had a broad increase in ME/CFS symptoms often lasting for several days, which was not seen in control subjects.[72] Patients who completed the Profile of Mood State questionnaire following CPET reported more fatigue, pain, and cognitive symptoms across the 3 days of data collection compared with values before CPET.[73] In contrast, Natelson's group, using actigraphy[74] and real-time report of symptoms onto a watch-like computer,[75] found a delay of 6 days before reduced activity and symptom worsening after CPET developed. Thus, these studies corroborated the occurrence of PEM following exertion, but noted a longer delay until objective evidence of its occurrence was seen. The variability in results across these studies suggests that PEM may vary both in intensity and time course—a point raised by the Institute of Medicine's report on ME/CFS.[3] One recent paper has noted that PEM is worse in patients with elevated lactate levels in the baseline condition (≥2 mmol/L);[76] confirming this result will be important. Another approach that has attracted more attention has sought to use exercise to precipitate PEM.

Using CPETs to Assess PEM

Two CPETs have been proposed as a technique by which to diagnose and quantify PEM in ME/CFS patients.[77–82] Maximal tests are performed on 2 consecutive days within a 24-hour period. Unfamiliarity with CPET testing and novelty inherent in doing this for the first time can lead to variability in the physiological measures obtained during this testing and improvement on repeat testing. However, this was not the case for ME/CFS patients, who were reported to show reduced maximal oxygen consumption (VO2max) and earlier onset of the anaerobic threshold, defined as the oxygen consumption at the ventilatory threshold (VO2VT). The authors inferred that these changes were a metabolic manifestation of PEM, but demonstrating a direct relation has not as yet been done.

With exercise, increased delivery of oxygen to the working skeletal muscle is needed. Cardiac output, lung function, hemoglobin concentration, vasodilatory capacity of the vasculature, and muscle metabolism are all key factors in determining oxygen availability. In healthy normal individuals, the major limitation to exercise performance is the ability to increase cardiac output, which occurs by heart rate acceleration, increase in contractility from catecholamine stimulation, and increase in stroke volume via the Frank Starling mechanism. In the absence of chronotropic incompetence, or heart failure, hypovolemia may impact cardiac output increase during exercise. Oxygen consumption at the anaerobic or ventilator threshold (VO2VT) is the point where carbon dioxide production increases relative to VO2, when lactate made via glycolysis (ie, anaerobic metabolism) is needed to meet metabolic demands that cannot be sustained by oxidative mechanisms (aerobic metabolism) alone.[83]

The reduction of VO2max and VO2VT on day 2 of 2-CPET studies in ME/CFS may be a biological marker for ME/CFS.[77,79] However, many of these studies lacked a comparable control group, which the investigators contended was not necessary because either of these variables do not decrease on repeat CPET in healthy people. However, the question remains whether sedentary, deconditioned control subjects would show the same response to 2-CPET as do more active healthy control subjects.

Although similar results with a decrease in VO2 max and VO2VT were found in subsequent studies,[80,81] more recent reports failed to find a reduction in day 2 VO2max but continued to find reductions in the VO2VT.[81,82] Most recently, a study from Norway showed that arterial lactate on day 2 CPET was increased at the VO2VT in comparison with day 1 CPET, whereas it was decreased in healthy control subjects[84]—a further confirmation of a metabolic abnormality occurring in ME/CFS. And another recent study shows that after day 2 CPET, female patients tended to show progressive disability that paralleled initial illness severity.[85] Some investigators[82] attributed the inconsistency of the data to differences in the patients' exercise intensity, ie, failing on day 2 to maximally exercise. More clinical research is needed to determine if serial cardiopulmonary exercise testing can provide an objective tool to assess PEM and the mechanism for exertion-induced induced PEM remains to be determined.[86]