Future Wellness Briefing · Issue 005 · Thursday 8 October 2026

Issue No. 005: One Number, Two Rooms

Two rooms sat within one degree of each other on the index occupational safety standards are written against. In one of them people sweated half as much again as in the other.

Plus: the strength case for ashwagandha, which rests on three trials and dissolves when the maths is corrected for how few trials that is.

The Future Wellness BriefingOne measured read a week. Free, and every position lands on the Register.

An index tells you the size of the problem. It hides which half of it you are meant to be controlling.

Two levers, one figure

Sweat doesn't cool you. Evaporating sweat cools you. The liquid takes heat from the skin in order to change into vapour, and how fast it can do that depends on how much water the surrounding air is already carrying. Dry air takes it readily. Saturated air won't take much at all, so the sweat runs off you and does almost nothing.

That leaves two separate levers in any warm room. Temperature sets how much heat has to go somewhere. Humidity sets how easily the main cooling route can carry it away. A heat index collapses both into one figure.

One degree apart on paper

Twenty adults walked at moderate intensity for two hours in each of two conditions: hot and dry at 40.6 degrees with 20 per cent humidity, then warm and humid at 30 degrees with 80 per cent humidity. On wet bulb globe temperature, the index those standards use, the two rooms came out at 29.3 and 28.3. The air temperature was more than ten degrees apart.

They sweated 0.53 litres an hour in the dry room against 0.35 in the humid one. That is half as much again, between two environments the index calls near enough identical.

The rule written against the index overwatered everybody. All of them drank to the standing occupational recommendation, 237 millilitres of water every fifteen to twenty minutes, aimed at keeping body mass loss under 2 per cent. Nobody came close to losing 2 per cent. Every participant finished heavier than they started, by 0.6 per cent in the dry room and 1.1 per cent in the humid one, carrying between 0.4 and 0.7 litres of surplus fluid. The authors' own word for the strategy is overprescribed.

Nothing flagged it. Not thirst, and not the hydration biomarkers. The only thing that separated the two rooms was what the body was doing in them.

The same split where people sleep

Across 236,115 nights from 1,067 users of a sleep app in South Korea, paired night by night with Seoul weather data, warmer nights came with less deep sleep, less total sleep, lower sleep efficiency and more time awake after falling asleep. Humidity tracked separately and differently, lowering sleep efficiency and lengthening wake time but hitting the stages in another pattern. Summer was the worst season.

Handle that one carefully. It reports directions and p-values rather than effect sizes, its sleep staging comes from a phone app rather than a sleep laboratory, and six of its ten authors work for the company that makes the app. It is directional support for a specification argument and not a dose.

The point: temperature and humidity are two variables and belong in a brief as two variables. A single index, a single set point or a single fixed rule gives you the total load and conceals the lever.

What to do this week

Separate the two numbers in any environmental brief. Where a thermal specification, a bedroom standard or a heat policy quotes one figure, ask what the dry bulb temperature was and what the humidity was, because the same index covers rooms the body treats very differently.

Take fixed-volume hydration out of your protocols. The volume that holds someone steady in dry heat puts them into surplus in humid heat, and in this trial it put them into surplus in both. Drinking to thirst is the alternative the authors raise, and it is cheaper to administer.

Put humidity control in the sleeping-room brief, alongside temperature, acoustics and light. The evidence for the bedroom is a good deal weaker than the evidence for the worksite, so make that case on thermal physiology and on comfort, not on a sleep-stage figure.

Evidence Check

Three trials deep

Ashwagandha is sold hard as a training adjunct. Somebody has now pooled only the trials that tested it that way, meaning randomised trials that combined a standardised extract with structured resistance training rather than the mixed bag of exercise contexts earlier reviews drew on.

What qualified. Four trials, of which three could be pooled, 161 people between them.

What it showed. Pooled conventionally, it beat placebo on bench press, on a lower-body lift and on chest girth, with nothing on body fat.

What happened next. The authors applied a correction built for pooling very few studies, and both strength intervals crossed zero. Removing any single trial moved each estimate again. GRADE certainty came back Low for the strength outcomes and Very Low for chest girth.

Why it generalises. Certainty in a pooled estimate is governed by how many trials there are rather than how many participants, so three studies and a thousand people is still three studies. Every trial here pointed the same way, which counts for something and isn't the same thing as a result that holds.

Our position: adjunctive ashwagandha for strength has not been shown. The larger trial base for this ingredient sits in anxiety, where fourteen randomised trials have been pooled in diagnosed patients. Verdict: sell it where it has been tested.

From the Academy

Reading an environment as its separate physical variables, rather than as the index that bundles them, is the built environment module of the Level 7 Diploma in Wellness Practice. Candidates are handed a room's measured conditions and marked on what they would control, in what order, and what they would refuse to promise. Applications for the next cohort open shortly. Read the prospectus.

Also this week

The ASA ran 309,000 paid-for wellness ads through its monitoring system and rated more than 86 per cent at medium or high compliance risk in every category it examined, mushroom supplements worst at 53 per cent high risk. This is AI-assisted classification rather than adjudication, so it measures prevalence and not enforcement. Prevalence at that level makes claims risk structural.
The ASA separately upheld both issues against an at-home blood testing service under CAP 3.1, 3.7 and 12.1, the first clean UK precedent on diagnostics-led wellness implying prevention and diagnosis.
Training older adults at their own measured peak hour built no more strength or lean mass than training them at their worst hour. Strength varied across the day by around 11 per cent on average, and by anywhere from 1 to 34 per cent between individuals, so the contrast was real. Aligning training to it changed nothing measurable. A gym selling circadian programming is selling adherence.
Three assigned protein targets, 0.8, 1.6 and 2.2 grams per kilogram a day, produced no detectable difference in muscle mass across twelve supervised weeks in 108 women aged 40 to 77. The training variable did move: concurrent training finished 2.04 percentage points lower on body fat, resistance training 6.68 kilos higher on leg press. Muscle mass was read by bioelectrical impedance, and a null doesn't establish equivalence.
A network meta-analysis of recovery modalities in elite endurance athletes found 35 of its 55 nodes resting on a single trial each. Sixteen trials, 273 participants, and of 47 comparisons against simply resting, three excluded the null and none robustly.
Oura postponed its Nasdaq listing on 29 September, having targeted a 15.62 billion dollar valuation, with indications that pricing would land at the low end of the range. Read it alongside Bain taking Vitabiotics at a reported 900 million pounds and P&G taking Thorne at a reported 3.8 billion dollars. Capital is paying for defensible science and traceable supply, and pricing consumer measurement on retention rather than on sensor accuracy.

Position

An index is a convenience for whoever writes the rule and a liability for whoever has to follow it. Ask for the underlying variables before you sign anything that quotes one.

For the sceptics

Fluid balance under two heat stresses. Randomised crossover, 20 healthy adults (10 female, 24 ± 5 y), two 120-min bouts of moderate-intensity intermittent walking (338 W) under hot-dry (40.6 C, 20% RH, WBGT 29.3 C) and warm-humid (30 C, 80% RH, WBGT 28.3 C) conditions, drinking 237 mL cold water every 20 min per NIOSH guidance. Whole-body sweat rate 0.53 ± 0.13 vs 0.35 ± 0.10 L/h (hot-dry vs warm-humid, P < 0.001); body mass gain 1.1 ± 0.5% (warm-humid) vs 0.6 ± 0.4% (hot-dry); net fluid balance 0.70 ± 0.27 vs 0.40 ± 0.26 L; hydration efficiency 48.9 ± 19.3% vs 28.2 ± 18.0% (all P < 0.001). Hydration biomarkers and thirst did not differ. No participant exceeded 2% body mass loss. No follow-up beyond the trial day. Paywalled. Zhao, X., Butts, C.L. and McKenna, Z.J. et al. (2026) Annals of Work Exposures and Health, 70(8). https://doi.org/10.1093/annweh/wxag085
Weather and sleep architecture. Observational analysis of 236,115 nights from 1,067 users of the SleepRoutine mobile application in South Korea over one year, each night paired with contemporaneous Seoul weather station data; nightly apnoea-hypopnoea index classified low (<15) or high (≥15). Elevated temperature associated with lower sleep efficiency, total sleep time and deep sleep ratio and with longer sleep latency and wake after sleep onset (all p < 0.001); higher humidity with lower sleep efficiency and light sleep ratio (p < 0.001, 0.004) and longer wake after sleep onset. Weather-associated changes were diminished on higher-AHI nights. Effect magnitudes not reported in the abstract; staging is app-derived rather than polysomnographic; six of ten authors are affiliated to Asleep Inc., which produces the application. Paywalled. Lee, M., Hong, H. and Kim, S. et al. (2026) Sleep and Breathing, 30(5). https://doi.org/10.1007/s11325-026-03824-7
Adjunctive ashwagandha and resistance-training adaptations. Systematic review and REML random-effects meta-analysis; four RCTs eligible, three poolable (n = 161), one preprint reported narratively; healthy adults in structured resistance training. 1-RM bench press SMD 0.60 (95% CI 0.20 to 1.01), lower-body 1-RM 0.52 (0.19 to 0.86), chest circumference 0.86 (0.41 to 1.31), body fat -0.32 (-0.72 to 0.08). I2 7 to 34%. Under Hartung-Knapp-Sidik-Jonkman adjustment both strength intervals crossed the null (bench -0.24 to 1.45; lower body -0.20 to 1.24), and each was fragile to single-trial removal. GRADE Low for both strength outcomes and body fat, Very Low for chest circumference. All three pooled trials raised at least some concerns on RoB 2. Review not registered; no external funding. Open access. Lee, J.-C. and Heo, A.-S. (2026) Nutrients, 18(17), 2815. https://doi.org/10.3390/nu18172815
Rhythm-aligned training in older adults. Three-arm, parallel-group, double-blind randomised controlled superiority trial; 108 randomised (41% female, mean age 67), 93 analysed; individual peak and trough performance times fixed by strength testing at 08:00, 12:00, 16:00 and 20:00 across four days; 12-week intervention, two supervised whole-body sessions weekly plus one 30-min endurance session. Primary outcome isometric midthigh pull 0.06 N/kg (95% CI -0.76 to 0.88); appendicular lean mass index 0.03 kg/m2 (-0.09 to 0.15). Mean diurnal variation in strength approximately 11%, individual amplitudes 1 to 34%. Prospectively registered, NCT06063135. No follow-up beyond the intervention. Open access. Bruggisser, F., Ritter, S. and Roth, R. et al. (2026) Journal of Cachexia, Sarcopenia and Muscle, 17(5). https://doi.org/10.1002/jcsm.70390
Protein target, training mode and muscle mass. Randomised 2 x 3 factorial trial; 108 women aged 40 to 77 assigned to 12 weeks of supervised resistance or concurrent training crossed with protein targets of 0.8, 1.6 or 2.2 g/kg/d; complete-case analyses included 83. Primary outcome BIA-derived skeletal muscle mass: no training x protein interaction (p = 0.856), no marginal protein effect (p = 0.726), no marginal training effect (p = 0.273). Concurrent training lower on body fat percentage by 2.04 points (95% CI -2.94 to -1.14); resistance training higher on leg-press estimated 1-RM by 6.68 kg (5.04 to 8.32); concurrent training higher on cycle-derived estimated VO2max by 4.53 mL/kg/min (3.80 to 5.25). Concurrent training carried greater total exercise exposure, so modality effects are not isolated. Null protein findings do not establish equivalence. Open access. Bagheri, R., Ghobadi, H. and Mansordehghan, M. et al. (2026) Journal of the International Society of Sports Nutrition, 23(1), 2720742. https://doi.org/10.1080/15502783.2026.2720742
Recovery modalities against passive recovery. Systematic review and random-effects network meta-analysis; 16 RCTs, 273 participants, 14 contributing to at least one network; competitive and elite endurance athletes; eight outcomes against passive recovery; CINeMA certainty against a minimal important SMD of 0.5; RoB 2. Of 47 contrasts, three excluded the null and none robustly; 35 of 55 nodes (64%) rested on a single trial. No modality outperformed passive recovery for delayed-onset muscle soreness (5 trials, I2 90.2%), creatine kinase (4 trials) or sport-specific performance (11 trials, I2 0%). PROSPERO CRD420261368621. Cited without an author list because none was returned on retrieval. Open access. (2026) Frontiers in Sports and Active Living, 8. https://doi.org/10.3389/fspor.2026.1876924
ASA compliance monitoring and the Menwell ruling. Regulatory material, not evidence. The ASA Pulse sweep analysed 309,000 paid-for online wellness ads across seven categories by AI-assisted classification and risk-rated them; no individual rulings were issued, so these are prevalence figures rather than enforcement. The Menwell adjudication upheld both issues under CAP 3.1, 3.7 and 12.1. Oura's IPO postponement and the Vitabiotics and Thorne transactions are company-sourced trade and business press, not independently audited here; several comparable deals in the same round-up are undisclosed.

Verification note: verified 8 October 2026 via PubMed, Scite and publisher records. Editorial notices were checked on every paper cited; none carries a retraction, correction or expression of concern. Two papers in the lead are paywalled and are reported from their abstracts. The weather and sleep paper carries a manufacturer affiliation, stated in the text above.

Written by Phil Learney. Every position we take is logged on the Register, with its evidence. Phil Learney is founder of HMN24, and the Briefing does not rule on HMN24 products.

Thursday mornings. One claim tested properly, what to do about it, and the position that goes on the Register.

It is the only wellness email I read where I can check the working.Estates director, five-property European hotel group

Every position lands on the Register

No brand is named in a negative finding, and where no trial exists we say never been tested rather than dressing up an absence as caution.