At some point on most health journeys, usually a few weeks in, a wearable appears on the shopping list. A smart ring, a watch, one of those screenless bands. The promise is appealing: objective numbers, nightly scores, a coach on your wrist.
The honest position, and the one this sector rarely gives you, is that a wearable in 2026 is two products in one. Part of it is genuinely excellent and independently validated. Part of it is confident-sounding guesswork. Knowing which part you are looking at, before you act on it, is the whole game.
The market in brief: watches, rings and bands
The choice now comes down to three form factors. Smartwatches remain the default, with the Apple Watch Series 11 from £369 in the UK carrying the broadest set of regulator-cleared health features and no subscription for any of them. Smart rings, led overwhelmingly by Oura, trade the screen for discretion and battery life. Screenless bands, which WHOOP defined, aim at people who want recovery data without another screen.
The bigger dividing line is the business model. Oura charges £5.99 a month or £69.99 a year and gates nearly everything beyond its three basic scores behind the membership. WHOOP sells no device at all, only annual memberships, which run £169 to £349 in the UK with the hardware included. Garmin and Polar take the opposite position: the full platform ships free with the device. Neither camp is wrong, but a ring or band whose useful features sit behind a fee should be budgeted as a recurring cost, not a one-off purchase.
If you are in the UK and already hold private health cover, check what your insurer offers before paying retail. Vitality currently offers the Apple Watch Series 11 from £9 upfront on 40-month interest-free credit, with the monthly payments falling as low as £0 depending on recorded activity.
What the evidence says they measure well
Independent validation research has now caught up with the marketing, and it is consistent about where the ground is solid.
Heart rate is the standout. A living systematic review of Apple Watch validation studies, covering 82 studies and 430,052 participants, found heart rate undershot by an average of just 0.27 beats per minute. A separate meta-analysis pooling 56 studies across brands put average heart rate error at 4.43% and step counting at 8.17%. For seeing whether you genuinely moved more this month than last, these numbers are dependable.
That same meta-analysis is equally clear about the weak spot: calorie burn estimates carried an average error of 27.96%, and the Apple Watch review rated energy expenditure error "inconsistent and frequently large". If your plan involves eating back the calories your watch says you burned, an error of that size can quietly erase your deficit while you follow every rule. The lead researcher's summary is the right frame for the whole device: "Think of it as a helpful guide, not a diagnostic tool."
Sleep scores: read the trend, not the verdict
Sleep tracking sits in between. A validation of six devices against clinical polysomnography, published in SLEEP Advances, found wearables detect sleep reliably (sensitivity 91.68% to 96.27%) but detect wakefulness poorly (specificity 29.39% to 52.15%). On sleep staging, the best performer managed only moderate agreement with the lab, and others only fair agreement.
Age matters too, and in an uncomfortable direction. A January 2026 study found the Oura Ring's total sleep time error grew from an average of 15.5 minutes in younger adults to 75.5 minutes in older adults, a meaningful gap given that older users are exactly who these devices' health claims court hardest.
None of this makes sleep tracking worthless. It makes the nightly number a rough sketch. If your sleep trend deteriorates for weeks, that is worth acting on; if last night scored 68 rather than 74, it is not. And a readiness, recovery or daily stress score deserves even less deference: no independent validation of any brand's proprietary scoring methodology has been published, for any brand.
Blood pressure: the feature to treat most carefully
This is the area where marketing and medicine have pulled furthest apart, and it matters because the stakes are real.
Apple's hypertension notification is the most credible entry: cleared by the FDA in September 2025 and trained on data from more than 100,000 participants. Yet an independent evaluation using US national health data found it detected only 41.2% of people with undiagnosed hypertension, against 92.3% specificity. The University of Utah researchers behind that work put it plainly: roughly 59% of people with undiagnosed high blood pressure would never receive an alert, and the feature should act as a supplement to, not a replacement for, "standard blood pressure screening with validated cuff-based devices".
The cardiology position is blunter still. A May 2026 editorial in the European Journal of Preventive Cardiology notes that "all recent hypertension guidelines explicitly recommend against the use of cuffless blood pressure devices for the management of hypertension", and warns that "regulatory clearance should not be conflated with independent validation of accuracy". The FDA itself issued a safety communication telling consumers not to rely on unauthorised blood-pressure-measuring wearables at all. A US clearance also has no legal standing in the UK.
The practical rule: a blood pressure alert from a wearable is a prompt to get a proper cuff measurement at a pharmacy or your GP surgery. Silence from your wearable is not an all-clear, and no wrist or finger device should be how you manage a diagnosed condition.
The cleared alerts with a better record are worth knowing about. Samsung's sleep apnoea notification was the first of its kind cleared by the FDA in 2024, irregular-rhythm notifications are cleared across Apple, Samsung and Garmin, and even a smart ring now carries FDA-cleared AFib detection. These have surfaced genuinely undiagnosed conditions. The same rule applies, though: an alert is a reason to see your GP, not a diagnosis, and symptoms such as breathlessness, chest pain or fainting warrant a GP appointment regardless of what any device says.
Using one well on a weight-loss journey
Many readers arrive at this question mid-journey, often on a GLP-1 medicine, and this is where a wearable earns its keep, used for the right job.
That job is consistency, not calories. Weight lost quickly is not all fat: a meaningful share is lean tissue, which is why resistance training and adequate protein matter throughout, something we cover in detail in our guides to exercise on Mounjaro and strength training on a GLP-1. A tracker cannot measure body composition, but its two reliable numbers, steps and heart rate, will tell you honestly whether the training habit is actually happening, week after week. That is valuable, because the habit is what protects the result.
The AI coaching layer is this year's big pitch: Google rebuilt Fitbit around a Gemini-powered health coach, WHOOP and Oura ship conversational assistants, and Oura has gone as far as a proprietary model for women's health. Treat these as what they are: articulate habit prompts built on the same imperfect data, not clinicians.
For progress itself, older instruments still win. A weekly weight trend, a tape measure and monthly photographs beat any wearable score, and our BMI calculator and weight loss calculator help set targets honestly, with the caveat that BMI says nothing about body composition.
Where the wearable stops helping
A wearable describes your whole system: total activity, whole-body weight, average sleep. It has nothing to say about the two problems that most often remain at the end of a successful journey. Rapid weight loss frequently leaves loose skin that no amount of tracking will tighten, and stubborn localised fat can outlast an otherwise finished plan. Localised treatments answer a different question from diet and activity: fat freezing, for instance, changes the shape of a treated area, and most patients see no meaningful change on the scale, which is precisely why no tracker will ever measure it.
If the number that bothers you has stopped being the one on the scale, that is the point where data ends and an honest conversation starts. A consultation at one of our clinics is an assessment of whether any treatment genuinely suits you, and a good one will tell you when the answer is no. VIVO Body Studio offers those consultations without obligation.
Where a wearable genuinely helps
- Heart rate and step counting are independently validated as accurate, which makes activity consistency easy to see and sustain
- Regulator-cleared alerts for irregular heart rhythm and sleep apnoea have surfaced real, undiagnosed conditions
- Worn consistently, week-on-week trends are meaningful even where the absolute numbers are not
Where it can mislead
- Calorie burn estimates carry an average error of around 28%, which can quietly erase a planned deficit
- Sleep staging agrees only moderately with sleep-lab measurement at best, and accuracy worsens with age
- Blood pressure features are explicitly discouraged by current hypertension guidelines, and readiness scores have no independent validation at all
Frequently Asked Questions
Can my smartwatch diagnose high blood pressure?
No. The most credible feature, Apple's FDA-cleared hypertension notification, is a screening prompt, not a diagnosis, and an independent evaluation found roughly 59% of people with undiagnosed hypertension would receive no alert. Current hypertension guidelines recommend against cuffless devices for managing blood pressure. Treat any alert as a reason to have a proper cuff measurement at a pharmacy or GP surgery, and never treat silence as an all-clear.
How accurate is sleep tracking on a wearable?
Good at telling that you slept, poor at telling how. In a six-device validation against clinical sleep measurement, wearables detected sleep with 91.68% to 96.27% sensitivity but detected wakefulness with only 29.39% to 52.15% specificity, and even the best device managed only moderate agreement on sleep stages. Accuracy also degrades with age. Use the trend, not the nightly score.
Do I need a subscription to get anything useful?
No. Oura and WHOOP gate their analytics behind memberships, but Garmin and Polar include their full platforms free with the device, and Apple requires no subscription for any of its core health features. The measurements that matter most on a weight-loss journey, steps, heart rate and activity trends, are free on most devices.
Is a wearable worth it while on Mounjaro or Wegovy?
It can be, for one specific job: making activity consistency visible. A meaningful share of weight lost on a GLP-1 medicine is lean tissue rather than fat, and resistance training plus adequate protein is the recognised way to protect muscle. A tracker will not measure body composition, but it will tell you honestly whether the training habit is actually happening.
Which numbers should I actually pay attention to?
Steps and active minutes for consistency, resting heart rate as a long-term trend, and cleared medical alerts as prompts to see a professional. For weight-loss progress itself, a weekly weight trend, a tape measure and monthly photographs will tell you more than any score your wearable produces.
