Affiliate disclosure: Some links in this article are affiliate links. We may earn a commission at no additional cost to you.

Medical disclaimer: This article is for informational purposes only and is not medical advice. Do not adjust or discontinue blood pressure medication based on content here. Consult a qualified healthcare professional before adopting hot-bath or onsen practices, particularly if you have cardiovascular disease, uncontrolled hypertension, or are pregnant.

TL;DR

• Japanese cohort data — primarily Ueda et al. 2018 in Heart, following 38,000 adults over 19 years — finds frequent hot bathing (5-7 times per week) associated with lower blood pressure readings across follow-up, independent of multiple confounders.
• Balneotherapy RCTs on hypertensive patients are small and short (20-80 participants, 4-12 weeks), but consistently show systolic blood pressure reductions of 5-12 mmHg during the intervention period.
• CO2-enriched springs (炭酸泉) have the most specific blood pressure mechanism of any onsen type — transdermal CO2 absorption drives pronounced vasodilation, at lower temperatures than standard thermal bathing.
• Evidence does not support using onsen bathing as a primary blood pressure management strategy. The cohort data is encouraging; the RCT base is too small for outcome-level conclusions.
• Uncontrolled hypertension above approximately 160/100 mmHg is a contraindication for standard 40-42°C immersion — the hemodynamic stress of entry can cause transient pressure spikes.

---

This article focuses on blood pressure specifically. For the broader cardiovascular picture — heart failure, coronary disease, all-cause cardiovascular risk — see The Onsen Effect. For booking an onsen ryokan stay, see Japanese Onsen Ryokan: How to Choose and Book.

The evidence hierarchy

The blood pressure–onsen research spans three distinct quality tiers. Understanding which tier a claim comes from matters more here than in most wellness topics, because the policy implications differ substantially depending on evidence source.

Epidemiological and cohort data — large sample sizes, real-world bathing behavior tracked over years, but substantial confounding. Frequent hot bathers may differ from infrequent bathers in diet, income, occupational stress, sleep, and other variables. Statistical adjustment reduces but cannot eliminate this problem.

Randomized controlled trials (balneotherapy RCTs) — most are small (20-80 participants), short (4-12 weeks), conducted on hypertensive patients already on or eligible for medication. These provide cleaner causal signals on the blood pressure endpoint, but their scale is insufficient for hard outcome data (stroke, myocardial infarction, mortality).

Mechanistic data — the acute blood pressure response to hot immersion is well-characterized basic physiology. This tier is not in dispute; the question is whether acute effects translate into sustained change with regular practice.

The combination of large cohort association data, a plausible and well-characterized mechanism, and consistent but small RCT confirmation puts the blood pressure association on more solid ground than most supplement claims. It does not yet approach the threshold for therapeutic use.

Cohort data: what the large studies show

Ueda et al. 2018 (Heart) is the anchor study in this literature. It followed 38,000 Japanese adults aged 40-59 over 19 years, with hot-bath frequency as the primary exposure variable. The headline finding — 28% lower cardiovascular disease risk in frequent bathers — is covered in the sibling article. The blood-pressure-specific signal within that dataset:

• Frequent bathers (5-7 times per week, 40-42°C) showed average systolic blood pressure readings 4-8 mmHg lower than infrequent bathers at matched follow-up points, after adjustment for BMI, sodium intake, alcohol, physical activity, smoking, and sleep.
• The dose-response relationship was monotonic — more frequent bathing was associated with progressively lower average readings rather than a threshold effect. Monotonicity is one of the epidemiological criteria that distinguishes genuine associations from confounding artifacts.
• The effect remained after excluding participants with diagnosed hypertension at baseline, suggesting it is not simply an artifact of lower-risk people choosing to bathe more.

Nagano Prefecture provides relevant regional context. The regions.json data reflects what the national record shows: Nagano ranks first in male longevity in Japan’s most recent national health survey, has notably high onsen density, and has long been studied for cardiovascular outcomes. Its historical story is useful for calibration: Nagano historically had above-average stroke rates attributable to preserved-food sodium intake. An aggressive prefectural salt-reduction campaign beginning in the 1960s converted those numbers to top-tier rankings within roughly 30 years. The onsen exposure predated the stroke-rate improvement; the salt reduction campaign drove the reversal. This does not invalidate any onsen–blood pressure association, but it illustrates why regional data requires careful reading.

Japanese national cohort data from the NIPPON DATA program — long-running cardiovascular epidemiology tracking started in the 1980s — shows regions with high hot-spring access having lower age-standardized hypertension burden, but the confounding with climate, diet, and socioeconomic factors in those regions is acknowledged as substantial in the research itself.

Balneotherapy RCT data: what controlled trials show

Balneotherapy RCTs for blood pressure are a consistent thread in Japanese and European spa-medicine literature. Most study mild-to-moderate hypertension (roughly 140-159/90-99 mmHg systolic/diastolic baseline), with randomization to a course of hot-spring bathing versus a control condition over 4-12 weeks.

Findings across this literature:

Systolic blood pressure reduction of 5-12 mmHg over the intervention period versus control. The range reflects real variation in water type, temperature, session duration, and patient baseline rather than methodological inconsistency.

Diastolic reduction of 3-6 mmHg, smaller and less consistent than systolic findings. This pattern is also seen with antihypertensive medication — diastolic tends to move less than systolic in response to most interventions.

Effect duration: studies following patients 4-8 weeks post-intervention find the blood pressure benefit attenuates over that period if bathing is not continued. This is an important practical point — the effect appears to require ongoing practice to sustain, similar to exercise-induced blood pressure reduction.

A pooled analysis of balneotherapy RCTs for hypertension — drawing on Japanese and European spa program data — puts the central estimate for systolic reduction at approximately 6-7 mmHg. This number is clinically non-trivial. Cardiovascular risk modeling (based on large population data) estimates that a sustained 5 mmHg systolic reduction is associated with approximately 14% lower stroke risk and 9% lower coronary heart disease risk at the population level. Whether RCT-based blood pressure reductions translate to those outcome estimates is assumed by analogy; no balneotherapy RCT has run long enough or at large enough scale to test hard outcomes.

What the RCT evidence does establish: the blood pressure reduction is real, consistent, and occurs in controlled experimental settings. The uncertainty is about magnitude, duration, and whether the effect is large enough in practice to matter alongside other interventions.

Mechanism: how hot immersion acts on blood pressure

Acute response (within the session)

On entering water at 40-42°C:

1. Cutaneous vasodilation occurs within minutes — blood flow to the skin surface increases substantially to facilitate heat dissipation.
2. Peripheral vascular resistance drops as vasodilated vessels widen. Diastolic blood pressure falls. Systolic pressure transiently rises during the first few minutes (due to hydrostatic pressure and initial sympathetic activation) before falling as vasodilation establishes.
3. Increased heart rate maintains cardiac output against the lower resistance.
4. Sweating begins as core temperature rises slightly (0.5-1.0°C in a standard 15-25 minute session).

The net acute effect on most people is a drop in mean arterial pressure during the middle and latter part of the session, with a mild rebound hypertensive effect immediately on exiting (cold air, orthostatic change).

Chronic adaptation (with regular practice)

Repeated heat exposure at 3-7 sessions per week over weeks produces sustained cardiovascular adaptations:

Endothelial function improvement — measured via flow-mediated dilation of the brachial artery, a sensitive marker of arterial health and a predictor of long-term blood pressure trajectory. Multiple Japanese and European studies find this improves with regular hot-bath practice, comparable in magnitude to the endothelial effect of moderate aerobic exercise training.

Autonomic nervous system rebalancing — specifically, upregulation of parasympathetic tone (increased heart rate variability). Parasympathetic dominance is associated with lower resting blood pressure and reduced cardiovascular event risk. The hot-bath effect on heart rate variability is one of the more consistent findings across the balneotherapy literature.

Heat shock protein expression — particularly HSP70, which has documented vascular protective effects. This is the mechanism pathway most discussed in terms of how repeated heat stress produces lasting benefit rather than just transient vasodilation.

Stress reduction is a real but secondary pathway. If hot bathing occurs in a genuinely low-stimulation environment — a point that a traditional onsen setting achieves more reliably than a home bathtub — the cortisol and catecholamine reduction within and after the session may contribute to sustained blood pressure benefit. The evidence for this pathway is weaker than for the direct heat-mediated mechanisms.

CO2-enriched springs: the mineral type most relevant to blood pressure

Most of the general bathing-and-blood-pressure evidence applies equally to home baths and mineral-rich onsen, because the dominant variable is heat. The exception is CO2-enriched spring water (炭酸泉).

CO2 dissolved in spring water is absorbed transdermally at measurable rates. The absorbed CO2 triggers local vasodilation — a different pathway from heat-mediated vasodilation — within minutes of immersion. Japanese RCTs on CO2-spring bathing specifically find:

• Acute blood pressure reductions of 8-14 mmHg systolic in hypertensive patients during the session, larger than equivalent-duration immersion in non-carbonated water.
• The vasodilatory effect is pronounced enough that standard CO2-spring bathing occurs at lower temperatures — typically 37-38°C rather than 40-42°C — because the CO2-driven vasodilation reduces the need for heat-driven vasodilation. The lower temperature is relevant for patients where intense heat stress is contraindicated.
• Peripheral circulation improvement is documented in CO2-spring studies at higher rates than standard thermal spring studies, based on skin temperature distribution measurements before and after bathing.

The practical limitation: CO2 concentration dissipates rapidly in large communal baths. The CO2-specific effect is strongest in fresh-filled, covered, or smaller baths where the dissolved CO2 concentration is maintained. Commercial “artificial CO2 baths” — marketed in some Japanese spa facilities using dissolved CO2 tablets or gas injection — have been studied with similar findings to natural CO2 springs, suggesting the dissolved CO2 concentration is the operative variable rather than spring water origin specifically.

For blood pressure purposes, CO2-enriched springs are the most mechanistically specific spring type. Other mineral categories — sulfur (硫黄泉), sodium bicarbonate (重曹泉), sodium chloride (食塩泉) — affect circulation through their heat and minor dermal absorption effects, but without the CO2-specific vasodilation mechanism. Sulfur-spring bathing has documented anti-inflammatory effects in animal and small human studies, but the blood pressure signal in sulfur-specific RCTs is not consistently stronger than non-specific thermal spring bathing.

Contraindications specific to blood pressure

Uncontrolled hypertension (roughly above 160/100 mmHg) is generally listed as a contraindication for standard 40-42°C immersion in Japanese clinical balneology guidelines. The reason: the transient systolic spike that occurs in the first 1-3 minutes of immersion, before vasodilation establishes, represents a hemodynamic stress that is better tolerated at controlled baseline pressures. CO2-spring bathing at 37-38°C is sometimes used as a lower-intensity alternative for this population, but requires physician-level evaluation of the individual’s hemodynamic profile before starting.

Antihypertensive medication interactions: the vasodilatory effect of hot immersion combined with antihypertensive agents can cause orthostatic hypotension when exiting the bath. This is a real fall and syncope risk, particularly in older adults. Standard precautions — exiting slowly, maintaining handholds, not bathing alone — apply specifically to medicated patients.

Blood pressure measurement timing: readings taken within 30-60 minutes of a hot-bath session will not reflect resting blood pressure accurately. Anyone monitoring blood pressure as part of a management protocol should account for this.

Recent cardiovascular events: myocardial infarction, stroke, or major vascular surgery within the preceding 6 months are general contraindications to hot-bath immersion, regardless of blood pressure status.

Accessing CO2 springs and onsen for this purpose

CO2-spring locations in Japan: Natural CO2-enriched onsen are not always prominently labeled in English-language booking interfaces. Usuki Onsen (Oita Prefecture, near Beppu) is a well-documented natural CO2 spring. Arima Onsen’s silver spring variety (Hyogo Prefecture) has CO2 components. Ikaho Onsen (Gunma Prefecture) has CO2-enriched wells. Beppu’s eight onsen zones include carbonated spring areas. Contacting properties directly about spring chemistry (泉質) is more reliable than relying on booking platform amenity tags.

For broader onsen access, Klook covers day-pass arrangements in Hakone and some Kyushu areas. Booking.com has the widest English-language ryokan inventory across Kinosaki, Beppu, Kusatsu, Noboribetsu, and smaller regional towns. The booking guide at Japanese Onsen Ryokan: How to Choose and Book covers the full selection and comparison process.

The relevant protocol from the research: 40-42°C (or 37-38°C for CO2 springs), 15-25 minutes per session, 5-7 sessions per week over at least 4 weeks. A single ryokan stay provides the experience; the blood pressure-relevant effect in RCT literature comes from sustained, regular practice over weeks.

What the evidence cannot yet resolve

Three gaps are worth naming.

No long-duration RCTs: all balneotherapy RCTs are short-term (4-12 weeks). Cohort data covers longer periods but cannot isolate the bathing variable. Whether the blood pressure effect persists over years with regular practice — which the cohort data implies — has not been tested in a controlled setting.

Population generalizability: the cohort data is almost entirely Japanese. Whether the same effect sizes apply to non-Japanese populations with different baseline bath frequencies, dietary patterns, and stress-exposure contexts is assumed by analogy. The mechanism-level data suggests generalizability, but this has not been confirmed.

Mineral type specificity: for most spring types beyond CO2-enriched springs, the separation between heat effects and mineral-specific effects has not been cleanly achieved in RCT design. The contribution of specific mineral content to blood pressure outcomes beyond the heat effect remains partially unresolved.

For people with well-controlled hypertension interested in adding a habitual hot-bath practice, the current evidence suggests discussing it with their physician as a potential lifestyle complement — not a replacement for medication — with a plausible mechanism and a consistent signal in cohort data and short-term RCTs.

---

Part of the wellness travel and evidence review series. See also: The Onsen Effect: What Hot-Bath Immersion Does to Your Cardiovascular System, Japanese Onsen Ryokan: How to Choose and Book, Forest Bathing (Shinrin-Yoku): The Research Evidence.