The Nervous System
Why your body never quite seems to switch off, even when you do. The autonomic story behind the second espresso, and what it costs to live in the always-on mode.
The series concludes hereYour body has two settings
One says go. The other says rest, repair, digest, breathe. Caffeine pushes hard on the first one and quietly turns down the volume on the second.
Your autonomic nervous system runs in the background. You don't think about it. It manages your heart rate, your breathing, your digestion, your blood pressure when you stand up, the dilation of your pupils, the sweat on your palms before a presentation. It does all of this without asking permission.
It runs on a balance between two branches. The sympathetic branch is your accelerator. The parasympathetic branch is your brake. A healthy nervous system flicks between the two fluently, hundreds of times an hour. You speed up when something matters. You slow down when nothing does.
Caffeine is a sympathetic activator. It nudges adrenaline and noradrenaline upward. It tightens blood vessels, raises blood pressure a touch, accelerates heart rate, sharpens attention. In the short term, useful. Across years of two or three cups a day, it tilts the system. The brake gets used less. The accelerator gets used more.
If you live with caffeine in your system most of the day, your body never gets a clean signal that it's safe to drop into rest. Sleep can still happen. Repair can still happen. But the depth and the quality of both quietly erode.
The number that tells you everything
It's called heart rate variability, or HRV. The clue is in the name. It's not how fast your heart beats. It's how variable the gap is between one beat and the next.
Counter-intuitively, more variability is better. A healthy nervous system produces beats that are slightly irregular, because your brake and your accelerator are constantly negotiating with each other in real time. A stressed or chronically sympathetic-dominant nervous system produces beats that are eerily regular, because the brake has gone quiet and the accelerator is calling the shots.
Your watch or ring measures this overnight. If the number is trending down over weeks and months while nothing else has changed, your nervous system is telling you something. Two of the most reliable things that pull HRV down are poor sleep and chronic caffeine load. Often the second is causing the first.
If your morning HRV has dropped and you don't know why, look at the cup
It is one of the most modifiable inputs you have. Cut caffeine after midday for two weeks and watch the number. If it climbs, you have your answer. If it doesn't, the cause is elsewhere and we can investigate.
HRV is not a vanity metric. It is a fairly honest readout of how well your body is recovering between days.
Three things that can change the picture quickly
When standing up shouldn't be hard
For some people, the nervous system isn't just nudged by caffeine. It's already struggling. If you're someone who feels lightheaded when you stand up, whose heart rate jumps thirty beats just from getting off the sofa, who feels worse rather than better after a hot shower, you may be sitting in the territory of dysautonomia.
This is more common than people realise, particularly in women, in people with hypermobility, in people with ADHD or autism, and in long COVID. POTS, which stands for postural orthostatic tachycardia syndrome, is one specific form of it. There are others.
Here's the relevant point. Caffeine has a complicated relationship with dysautonomia. In small, well-timed doses some people genuinely feel better, because caffeine increases vascular tone and helps blood return from the legs to the head. In larger or chronic doses it makes things worse, because it adds sympathetic load on top of a system that is already firing too hard.
If standing makes you dizzy, if your heart races at rest, if you have to sit down in queues, please bring this to a clinician. Don't manage it with stronger coffee. There are proper assessments and treatments available, and getting them right matters more than any caffeine adjustment will.
Steadying yourself on the chair every time you stand isn't normal
Many people have lived with this for years and assumed it's just how they are. It isn't. It's a treatable pattern, and identifying caffeine's role in it, helpful or unhelpful, is part of the assessment.
The point isn't that caffeine is good or bad here. The point is that for nervous systems running close to the edge, caffeine's effect is not neutral and it deserves attention.
Your nervous system is not a battery. It is a balance. Caffeine helps you find the gear up. It rarely helps you find the gear down.
Resetting the brake
The parasympathetic branch responds to specific things. None of them are dramatic. None of them require an app. All of them ask for the same currency: a few quiet minutes, repeated.
Five things that genuinely help
- Slow exhalations. A longer out-breath than in-breath, for a few minutes. Five seconds in, eight seconds out. The vagus nerve is listening.
- Gentle warmth on the face. A warm flannel, a warm shower, sunlight on closed eyes. Not heat. Warmth.
- Walking without your phone. Twenty minutes. The point isn't the steps. The point is that nothing is asking anything of your attention.
- Eating without doing anything else. No phone, no screen, no podcast. The body recognises that meal as a parasympathetic event. Digestion improves. So does sleep that night.
- A morning without coffee. Not forever. Once a week. See what happens.
Vagal tone is trainable. People who breathe slowly, walk regularly, sleep enough and eat without scrolling have measurably higher HRV after a few weeks. None of this is mysterious. It's just rarely prescribed because it can't be sold.
Mechanism
Caffeine antagonises adenosine A1 and A2A receptors centrally and peripherally. The peripheral effect on adenosine A2A in vascular smooth muscle and on A1 in cardiac tissue produces transient pressor effects, modest tachycardia and increased circulating catecholamines, particularly noradrenaline. Mean blood pressure rises 3 to 8 mmHg systolic acutely. Heart rate response is more variable and depends on baroreflex sensitivity.
More relevant for chronic exposure is the effect on autonomic balance. Time-domain HRV markers (RMSSD, SDNN) and frequency-domain markers (HF power, LF/HF ratio) shift consistently in the direction of sympathetic predominance with regular caffeine intake, particularly in slow metabolisers (CYP1A2 *1F homozygotes). The shift is dose-dependent and partially reversible within 14 days of cessation in most patients.
The clinical question is not whether caffeine affects autonomic tone. It clearly does. The question is whether that effect matters for the patient in front of you. In many patients it does not. In a meaningful subset it does, and identifying that subset is the work.
Trending HRV decline on consumer wearables (Whoop, Oura, Apple Watch, Garmin) without other obvious lifestyle change.
Resting heart rate creep of 5 to 10 bpm over weeks to months.
Orthostatic symptoms with tachycardia on active stand of 30+ bpm sustained at 10 minutes (or 40+ bpm in adolescents).
The neurodivergent and hypermobile populations
Dysautonomia prevalence is substantially elevated in autism, ADHD, hypermobile EDS and hypermobility spectrum disorder. The mechanisms differ but converge functionally on baroreflex dysfunction, abnormal venous return and increased sympathetic drive at rest. POTS in particular has a strong association with hEDS and with female sex.
For these patients, caffeine sits in an unusual clinical space. Acutely it can improve symptoms by increasing peripheral vascular resistance and supporting cerebral perfusion on standing. This is why some POTS patients describe themselves as feeling worse without it. Chronically and at higher doses, the same patients often deteriorate, with worsening sleep, increased palpitations, exacerbated tachycardia and reduced HRV.
The threshold dose at which the curve inverts is highly individual. A reasonable starting framework is to think of caffeine in dysautonomic patients the way you might think of a short-acting stimulant: useful as a positioned intervention, problematic as a chronic baseline.
HRV as a clinical signal
Consumer HRV data is imperfect but useful, particularly when patients have weeks or months of baseline. Look for trends rather than absolute values. Population norms are wide and individual baselines vary substantially with age, sex and fitness. What matters is movement away from the patient's own baseline.
A reproducible 15 to 25 percent decline in nightly RMSSD in a previously stable adult, in the absence of acute illness, training load change, alcohol increase or major life stressor, is clinically interesting. Caffeine intake (timing more than dose) is one of the highest-yield modifiable inputs to test.
Practically: ask the patient to keep their dose constant but stop all caffeine after midday for 14 days. The HRV signal usually moves within 7 to 10 days. If it moves, the lever is identified. If it doesn't, look elsewhere (sleep apnoea, alcohol, subclinical thyroid disease, depression, perimenopause).
What's worth measuring, what isn't
RMSSD overnight is the most clinically useful HRV metric on consumer devices. It reflects parasympathetic activity reasonably well and is less artefact-prone than spot readings. Frequency-domain metrics (HF, LF/HF) are useful when available but more fragile.
Resting heart rate trend lines are also valuable and often easier for patients to engage with. A creeping resting heart rate is a quiet but real signal.
Differential considerations
Always exclude
Subclinical hyperthyroidism, anaemia, sleep-disordered breathing, alcohol load, supplements containing yohimbine or synephrine, and medication interactions (particularly SSRIs, SNRIs and atomoxetine).
Often missed
Perimenopausal autonomic shift in women aged 40 to 55. Hypermobility-associated dysautonomia in young adults presenting with anxiety. Iron deficiency without anaemia in menstruating patients.
Patients on methylphenidate, lisdexamfetamine or atomoxetine have a different ceiling. Adding 200 to 400mg of caffeine on top of therapeutic stimulant doses meaningfully shifts the autonomic profile, raises blood pressure and heart rate, and degrades sleep architecture. This is not always obvious to patients, who often describe themselves as being on "just a normal amount of coffee".
Active stand test, in clinic, takes ten minutes
Lie supine for five minutes, record heart rate and blood pressure. Stand. Record at one, three, five and ten minutes. A sustained rise in heart rate of 30 bpm or more (40 in adolescents) without orthostatic hypotension supports POTS. A drop in systolic of 20 or diastolic of 10 supports orthostatic hypotension.
It is one of the highest-yield investigations available in a primary care or psychiatric outpatient setting and it requires no equipment beyond a stopwatch and a sphygmomanometer.
In autonomically vulnerable patients, caffeine isn't a lifestyle question. It's a dose-response intervention with measurable downstream consequences.
Restoration: what actually moves the needle
Vagal tone is trainable. Slow paced breathing at approximately six breaths per minute (4-second inhale, 6-second exhale) for ten minutes daily produces measurable HRV gains across two to four weeks in most patients. Cold-water face immersion (mammalian dive reflex) produces transient parasympathetic activation and is occasionally useful as an acute intervention for tachycardia.
Aerobic exercise improves baroreflex sensitivity. Resistance training increases venous return capacity and is particularly useful in hypermobile patients. Sleep extension by 60 to 90 minutes nightly produces consistent HRV gains. None of these are exotic. All require sustained patient engagement, which is the actual clinical challenge.
Patients with autonomic dysfunction are often already overwhelmed. A single targeted intervention, well-explained and reviewed at four weeks, outperforms a comprehensive lifestyle plan that is never started. Caffeine timing change is often the highest-yield first step.
When to refer
- Sustained orthostatic tachycardia of 30+ bpm with symptoms despite caffeine and lifestyle modification.
- Syncope, presyncope of any frequency, or falls related to postural change.
- Suspected hEDS or hypermobility spectrum disorder with autonomic features (Beighton score, family history).
- Resting heart rate persistently above 100 bpm in the absence of obvious cause.
- HRV decline of 30%+ from baseline that does not recover with intervention.
- Any patient with POTS-like features who is currently using high-dose caffeine or stimulants without specialist input.
Syncope with exertion, syncope with palpitations, family history of sudden cardiac death, or any new significant arrhythmia warrants cardiology referral irrespective of caffeine use. Do not assume autonomic dysfunction without ruling out structural and electrical cardiac causes first.
The cup was always more interesting than we treated it as
Across nine sections we've looked at caffeine through mechanism, attention, sleep, stress, mood, hormones, the body, hidden reactions and now the nervous system itself. The aim was never to tell you to stop drinking coffee. It was to give you enough understanding to make the decision yourself, and to recognise when something the cup is doing might deserve a clinical conversation.
If anything in this series has resonated, that's worth following up.
Get in touch with Sanctum
