Autonomic reactivity to stress in multiple sclerosis: A meta-analytic review.

This meta-analysis examined whether people with multiple sclerosis (MS) show blunted autonomic nervous system responses to emotional or psychological stress compared to healthy controls. The researchers pooled data from 14 studies that measured:

**Heart rate (HR)** — beats per minute during stress versus baseline

**Heart rate variability (HRV)** — specifically RMSSD (root mean square of successive differences), HF power (high-frequency power), and LF/HF ratio (low-frequency to high-frequency ratio), which reflect the balance between sympathetic and parasympathetic activity

**Blood pressure (BP)** — systolic blood pressure changes during stress

**Electrodermal activity (EDA)** — skin conductance level and skin conductance responses, reflecting sweat gland activity controlled by the sympathetic nervous system

**Pupil diameter** — changes in pupil size during emotional stimuli

The comparators were healthy control groups (people without MS) who underwent the same stress-inducing tasks. The stress tasks included things like watching emotionally charged images, performing mental arithmetic under time pressure, giving a speech, or solving difficult cognitive tests.

The meta-analysis included 14 studies with a total of 372 participants with multiple sclerosis and 292 healthy controls. Specific sample sizes per study ranged from 12 to 60 participants per group. The MS participants included people with relapsing-remitting MS, secondary progressive MS, and primary progressive MS. Most studies recruited from hospital neurology departments or MS clinics. Ages ranged from approximately 30 to 55 years. Disease duration varied across studies, typically 5–15 years since diagnosis. Exclusion criteria common across studies included: other neurological disorders, cardiovascular disease, diabetes, use of medications affecting autonomic function (beta-blockers, anticholinergics), and current psychiatric disorders other than anxiety or depression.

The studies used a combination of physiological recording equipment and standardised stress-induction protocols:

**Heart rate and HRV:** Measured using electrocardiography (ECG) with electrodes placed on the chest. HRV metrics were calculated from R-R intervals (the time between heartbeats). RMSSD reflects parasympathetic activity; HF power reflects vagal tone; LF/HF ratio reflects sympatho-vagal balance.

**Blood pressure:** Measured using continuous finger cuff monitors (e.g., Finapres) or standard arm-cuff sphygmomanometers at fixed intervals during stress tasks.

**Electrodermal activity:** Measured using electrodes placed on the palm or fingers, recording skin conductance level (tonic) and skin conductance responses (phasic) to specific stimuli.

**Pupil diameter:** Measured using eye-tracking cameras that record pupil size changes in millimetres.

**Stress-induction tasks:** Included the International Affective Picture System (IAPS — standardised emotional images), the Trier Social Stress Test (speech and mental arithmetic in front of an audience), the Stroop colour-word interference task, and mental arithmetic under time pressure with negative feedback.

**Study design:** This is a meta-analysis and systematic review following PRISMA guidelines. The authors searched PubMed, Scopus, and Web of Science from database inception to February 11, 2025. Two independent reviewers screened abstracts and extracted data. Fourteen studies met inclusion criteria.

**Inclusion criteria:** (1) adult participants (≥18 years) with MS; (2) comparison with healthy controls or correlations with MS symptom severity; (3) no major neurological comorbidities; (4) detailed statistical data reported; (5) published in English; (6) assessment of autonomic responses; (7) inclusion of emotional or stress-inducing tasks.

**Statistical approach:** The meta-analysis used a random-effects model, which assumes that true effect sizes vary across studies (appropriate when studies differ in methods, populations, and stress tasks). Effect sizes were calculated as Cohen's d (standardised mean difference). For between-group comparisons (MS vs. controls), the pooled effect size was based on 6 studies that reported cardiovascular data (HR, HRV, BP). Heterogeneity was assessed using the Q statistic, I² index (percentage of variation due to true heterogeneity rather than chance), and τ² (between-study variance). Publication bias was assessed using Egger's regression test and fail-safe N.

**What this design can and cannot prove:**

*Can prove:* That people with MS, on average across studies, show a different magnitude of autonomic response to stress compared to healthy controls. The meta-analytic approach increases statistical power and generalisability by combining multiple studies.

*Cannot prove:* Causality — the blunted response could be due to MS-related brain damage, but could also reflect medication effects, deconditioning, or differences in how people with MS perceive or appraise stress. The meta-analysis cannot determine whether blunted autonomic reactivity causes emotional disturbances or is a consequence of them. It also cannot identify which specific brain lesions are responsible.

**Major methodological weaknesses:**

**High heterogeneity (I² = 93.6%):** This is extremely high, meaning the studies varied enormously in their results. Some studies showed large blunting, others showed little difference. This makes the pooled estimate less reliable.

**Only 6 studies in the main meta-analysis:** A small number of studies limits statistical power and generalisability.

**Varied stress tasks:** Different studies used different types of stress (emotional images vs. social stress vs. cognitive stress), which may engage different neural circuits.

**Medication confounds:** Many MS medications (e.g., interferon-beta, fingolimod) can affect heart rate and autonomic function, but most studies did not adequately control for this.

**Publication bias:** The authors did not report Egger's test results in the abstract, and with only 6 studies, publication bias assessment is unreliable.

**Qualitative synthesis for EDA and pupil data:** These measures were not included in the meta-analysis, so conclusions about them are weaker.

**Primary outcome — Between-group comparison (MS vs. controls):**

Pooled effect size across 6 studies: d = −1.21, 95% CI [−2.23, −0.19], p = .020

This is a **large** effect size, indicating that people with MS show substantially blunted autonomic responses to stress compared to healthy controls

Heterogeneity: Q = 47.34, p < .001; I² = 93.6% (very high); τ² = 1.50

**Secondary outcome — Within-group comparison (MS only):**

Within-group analysis showed a moderate increase in autonomic responsiveness during stress compared to baseline: d = 0.71, p = .079

This was **not statistically significant**, suggesting that while people with MS do show some autonomic response to stress, it is reduced and inconsistent

**Qualitative findings from other autonomic measures:**

Electrodermal activity: Most studies reported reduced skin conductance responses in MS compared to controls during emotional stimuli

Pupil diameter: Limited data, but one study found reduced pupil dilation to emotional images in MS

Blood pressure: Some studies found blunted systolic BP responses to stress in MS, others found no difference

**Publication bias:**

Fail-safe N was computed to determine how many null-result studies would be needed to invalidate the findings (specific number not reported in abstract)

Egger's regression test was performed (results not reported in abstract)

The effect size of d = −1.21 means that the average autonomic response in people with MS was about 1.2 standard deviations lower than in healthy controls. To put this in concrete terms:

In healthy controls, heart rate typically increases by 10–15 beats per minute during a stressful task like public speaking

In people with MS, the increase might be only 2–5 beats per minute — roughly a 60–80% reduction in the normal stress response

For heart rate variability (RMSSD), which normally decreases during stress as the body shifts toward sympathetic dominance, people with MS showed a much smaller decrease, suggesting their nervous system is less able to shift gears in response to challenge

This blunting is comparable in magnitude to the effect of taking a beta-blocker medication (like propranolol) before a stressful event, or to the reduced stress reactivity seen in people with chronic fatigue syndrome or depression.

**Acknowledged by authors:**

Small number of studies available for meta-analysis (k = 6)

High heterogeneity across studies, limiting confidence in the pooled estimate

Limited data on non-cardiovascular autonomic measures (EDA, pupil response)

Variability in stress-induction methods across studies

Potential confounding by MS medications that affect autonomic function

Lack of standardisation in how autonomic reactivity was quantified

**Critical reader observations:**

**Medication confounds are severe:** Many MS disease-modifying therapies (fingolimod, siponimod, ozanimod) directly affect heart rate by acting on sphingosine-1-phosphate receptors. Interferons can cause flu-like symptoms that alter baseline autonomic state. Most studies did not adequately control for or report medication status.

**Disease subtype and severity not analysed:** The meta-analysis did not separate relapsing-remitting from progressive MS, or account for disability level (EDSS scores). Blunted reactivity might be more pronounced in progressive forms or in people with higher disability.

**No analysis of lesion location:** MS lesions in the brainstem, hypothalamus, or insula would be expected to affect autonomic responses, but no study included neuroimaging correlates.

**Publication bias likely:** With only 6 studies, the meta-analysis is vulnerable to the "file drawer problem" — studies finding no difference may never have been published.

**No correction for multiple comparisons:** The authors conducted multiple analyses (between-group, within-group, multiple autonomic measures) without adjusting significance thresholds.

**Emotional state not controlled:** People with MS have higher rates of anxiety and depression, which themselves alter autonomic reactivity. The blunted response could reflect comorbid depression rather than MS-specific damage.

**Resting baseline differences:** If people with MS already have altered resting autonomic function (e.g., lower HRV at baseline), then stress-induced changes may appear blunted simply because they start from a different baseline.

For someone running their own n=1 experiment (whether you have MS or not, and want to understand your stress reactivity):

### What to test

**Specific intervention:** Heart rate variability biofeedback training (also called resonance breathing or paced breathing at ~6 breaths per minute). This is a well-studied technique that increases vagal tone and may improve stress reactivity. Alternatively, test cold exposure (e.g., cold showers or ice baths) which activates the sympathetic nervous system and may train autonomic flexibility.

**Dose:** For HRV biofeedback, practice 20 minutes daily of paced breathing at 5.5–6.5 breaths per minute (inhale 4–5 seconds, exhale 5–6 seconds). For cold exposure, 2–3 minutes of cold water (10–15°C) once daily.

### Minimum meaningful duration

**HRV biofeedback:** At least 4 weeks of daily practice. Studies show measurable changes in resting HRV and stress reactivity after 4–8 weeks.

**Cold exposure:** At least 2–3 weeks to see adaptation effects. Acute effects are immediate but habituation takes time.

### What to measure

**Primary metric:** Heart rate variability — specifically RMSSD (reflects parasympathetic activity) and LF/HF ratio (reflects sympatho-vagal balance). Measure at the same time each day, in the same posture, after 5 minutes of quiet rest.

**Stress reactivity test:** Measure HRV and heart rate before, during, and after a standardised stressor. A simple option: 2 minutes of mental arithmetic (count backwards by 7 from 1000 as fast as possible while being timed). Record HR and HRV for 2 minutes before, 2 minutes during, and 2 minutes after recovery.

**Secondary metrics:** Subjective stress rating (1–10 scale before and after the stressor), and a simple reaction time test (available on phone apps) to measure cognitive performance under stress.

### Key confounds to control for

**Time of day:** Measure at the same time daily. HRV follows a circadian rhythm (typically highest in the morning).

**Caffeine and alcohol:** Avoid for 4 hours before measurement. Caffeine increases sympathetic activity; alcohol reduces HRV.

**Meals:** Measure at least 2 hours after eating. Digestion alters autonomic balance.

**Medications:** If you take beta-blockers, anticholinergics, or MS disease-modifying therapies (especially fingolimod), these will directly affect HRV. Document your medication timing and consider whether you can safely test on medication-free days (consult your doctor first).

**Menstrual cycle:** If applicable, HRV varies across the menstrual cycle. Track cycle phase and compare measurements within the same phase.

**Physical activity:** Avoid vigorous exercise for 2 hours before measurement. Exercise acutely increases sympathetic activity.

**Sleep:** Poor sleep reduces HRV. Track sleep quality and duration.

### What a positive result would look like

**For HRV biofeedback:** After 4 weeks, you should see a 10–20% increase in resting RMSSD (e.g., from 30 ms to 36 ms). During the stress reactivity test, your heart rate increase should be smaller (e.g., from +15 bpm to +8 bpm), and your HRV should recover to baseline faster (within 2 minutes instead of 5 minutes).

**For cold exposure:** After 2–3 weeks, you should see a smaller heart rate increase during cold exposure (e.g., from +20 bpm to +10 bpm), and faster return to baseline after exiting the cold. Your subjective cold discomfort rating should decrease by at least 2 points on a 10-point scale.

**A negative result** (no change in HRV or stress reactivity) could mean: (a) the intervention dose was too low or too short, (b) you have a medical condition or medication that blunts autonomic responses (as this meta-analysis suggests for MS), or (c) your baseline HRV is already optimal and has limited room for improvement.

**Important caveat for people with MS:** This meta-analysis suggests that if you have MS, your autonomic responses may be inherently blunted due to central nervous system damage. This does NOT mean interventions are useless — it means you may need longer training periods (8–12 weeks instead of 4 weeks) and should expect smaller absolute changes. It also means that subjective stress ratings may not match physiological measures — you might feel stressed even though your heart rate barely changes. Track both subjective and objective measures to understand your personal pattern.