Weight loss didn't improve AF symptoms in older adults — here's what that means for dietary advice

A rigorous UK randomized trial published May 20, 2026 in JAMA reached an uncomfortable conclusion: dietary weight loss produces no meaningful improvement in atrial fibrillation (AF) symptoms or arrhythmia burden in older adults with established persistent AF — even when participants successfully lost nearly 10% of their body weight. The finding matters not because it condemns weight management, but because it forces a more precise answer to a question clinicians and patients have been asking for years: does losing weight actually fix an irregular heartbeat?

The short answer from LOSE-AF: not in older patients who already have persistent AF, and not at a ~10% reduction. The more useful answer is about timing, magnitude, and realistic expectations — which this article translates into a concrete dietary recommendation below.

Study type: Randomized controlled trial (RCT)

pubmed.ncbi.nlm.nih.gov

正在加载链接预览…

Who was studied and what was the intervention

The LOSE-AF (Weight Loss in Older Patients With Persistent Atrial Fibrillation) trial 1 enrolled 118 adults across two UK hospitals between November 2018 and April 2025. All participants were aged 60–85 (mean 68, SD 6), had a body mass index of at least 27, and had been referred for electrical cardioversion due to persistent AF. Approximately one-third (33%) were women.

Participants were randomly assigned 1:1 to one of two groups:

Intervention (n = 59): An 8-month structured program combining a low-calorie diet with individualized behavioral support. Over 32 weeks, participants received regular scheduled appointments covering behavioral coaching, weight monitoring, and access to formula meal-replacement products.
Control (n = 59): Standard care — a single consultation with a research nurse plus written materials on weight-loss dieting.

The trial was open-label (neither participants nor researchers could be blinded to group assignment), which is unavoidable in a lifestyle intervention study. The primary outcome was change in the AF Severity Scale (AFSS) symptom severity subscale — a validated 0–35 instrument where higher scores indicate more severe AF symptoms — measured at 8 months.

The trial was funded by the NIHR Oxford Biomedical Research Centre and the British Heart Foundation. 1 The study ran across nearly seven years in part due to interruptions from the COVID-19 pandemic.

The primary result: no change in AF symptoms

After 8 months, AFSS symptom severity scores did not differ meaningfully between groups. 1

The intervention group scored 7.9 (SE 0.84) on the AFSS subscale; the control group scored 8.9 (SE 0.84). The between-group difference was −0.9 points (95% CI −3.3 to 1.4; P = 0.43) — statistically null, and too small to constitute a clinically meaningful change on a 0–35 scale.

"Our study demonstrates that weight loss in older adults with AF is achievable and safe, but it does not translate into meaningful improvements in arrhythmia burden or symptom relief."
— Dr. Matteo Sclafani, lead author (Oxford University Radcliffe Department of Medicine) 2

Pre-specified subgroup analyses by sex, age, and baseline BMI found no evidence that the intervention benefited any particular subgroup. 1 Every secondary AF outcome — arrhythmia burden measured by a 2-week wearable cardiac monitor, rate of repeat cardioversion, and rate of AF ablation — also showed no statistically significant difference between arms. 1

Weight loss was real and durable — AF improvement was not

The null result is not explained by a failed intervention. The dietary program did what it was designed to do.

Outcome	Intervention group	Control group	Difference (95% CI)	P value
Adjusted weight at 8 months	92.6 kg (SE 0.85)	99.4 kg (SE 0.85)	−6.9 kg (−9.2 to −4.5)	<0.001
% body weight lost	9.7%	3.1%	—	<0.001
AFSS symptom severity score	7.9 (SE 0.84)	8.9 (SE 0.84)	−0.9 (−3.3 to 1.4)	0.43
AF burden (wearable monitor)	No significant difference	—	—	NS
Physical performance	No significant difference	—	—	NS
Blood pressure, lipids, cardiac imaging	No significant differences across any measure	—	—	NS

1 2

The weight difference persisted at a median 3.5-year follow-up — a longer horizon than most dietary intervention trials sustain. 2 Neither group experienced serious adverse events related to trial participation. Notably, physical performance did not decline in the intervention arm — a reassuring safety signal given that weight loss in older adults can sometimes accelerate muscle loss.

As co-investigator Dr. Rohan Wijesurendra (Oxford Population Health) stated: "These findings suggest that while weight loss has general health benefits, additional therapies are needed to effectively manage AF in older adults." 2

cardiologynownews.org

正在加载链接预览…

Why LOSE-AF and an earlier trial reached opposite conclusions

The finding is newsworthy precisely because it contradicts the LEGACY trial — the only prior RCT to show that weight loss reduces AF severity and burden. Understanding why the two trials diverged is essential before adjusting clinical guidance.

The LEGACY trial 3 enrolled approximately 150 participants with a mean age of around 60 and achieved a substantially larger weight reduction — participants in the intervention arm dropped from a BMI of roughly 32.8 to 27.2, roughly a 17% reduction in body weight. LEGACY was conducted at a single center at the University of Adelaide by the research group of Prof. Prashanthan Sanders.

LOSE-AF differed on three dimensions that, taken together, may explain the gap:

Older, longer-established disease: LOSE-AF participants averaged 68 years old (vs. ~60 in LEGACY) and had persistent rather than paroxysmal AF — a pattern less likely to respond to single-factor interventions once the underlying substrate is established.
Smaller weight reduction: 9.7% vs. 17%. Whether a threshold effect exists near 15% weight loss — below which AF rhythm benefits do not appear — is biologically plausible given the dose-response pattern, but has not been confirmed by a third RCT.
Independent replication: Virtually all prior RCT evidence linking weight loss to AF improvement came from the same Adelaide research group. 3 LOSE-AF was the first independent attempt to reproduce those findings — and it did not.

Gregory M. Marcus (MD, MAS), vice-dean at UCSF and associate editor at JAMA, wrote the accompanying editorial and identified the weight loss magnitude and the prevalence of longstanding persistent AF as the two most likely explanatory factors. 3 He also called attention to the broader evidentiary context: obesity plausibly promotes AF via sleep apnea, systemic inflammation, hypertension, and left atrial enlargement — but in patients who already carry persistent AF, reversing one upstream risk factor may not be sufficient to reverse a disease process already in progress.

www.researchgate.net

正在加载链接预览…

Limitations

Several limitations constrain how broadly LOSE-AF's null result can be applied.

Open-label design: Participants knew their group assignment. Symptom reporting on a subjective scale like the AFSS may be influenced by awareness of receiving a more intensive intervention (placebo effect in reverse). This is an inherent limitation of all lifestyle RCTs and cannot be eliminated.

Sample size: 118 participants is sufficient for the power calculation the authors pre-specified, but modest enough that small true effects, if they exist, might not be detected.

Single-country, two-hospital sample: The UK healthcare context, the specific commercial meal-replacement products used, and the demographics of participants referred for cardioversion in England may not generalize to North American or East Asian populations with different baseline diets, BMIs, or AF management pathways. 1

AF type: All participants had persistent AF — meaning the arrhythmia continues without interruption, rather than coming and going in discrete episodes. The intervention might produce different results in paroxysmal (intermittent) AF, where the arrhythmic electrical substrate is less structurally entrenched. 4

Incomplete secondary data: Some extended rhythm monitoring and cardiac imaging assessments were not completed for all participants, in part due to COVID-19 disruptions. The precise effect estimates on secondary endpoints should be interpreted with that caveat. 4

The dietary takeaway — a concrete recommendation

LOSE-AF does not argue that weight loss is useless for patients with AF. It argues that the dosing window, the disease stage, and the therapeutic expectation need to be recalibrated.

For health-conscious individuals managing cardiovascular risk:

Weight loss remains a well-supported strategy for reducing the development of AF and for improving overall cardiovascular health — blood pressure, lipid profiles, diabetes risk, and left atrial remodeling. LOSE-AF enrolled patients who already had established persistent AF; the null result applies to that specific population and outcome, not to prevention.

For older adults with diagnosed persistent AF:

A low-calorie dietary intervention can safely produce meaningful weight loss (~10%) without increasing fall risk, frailty, or physical decline — the trial found no adverse effect on physical performance. That is itself a clinically useful finding: older patients with AF who want to lose weight for general cardiovascular health can do so without the concern that dietary weight loss will worsen their functional status.

What dietary weight loss alone should not be positioned as in this population is a primary strategy for AF rhythm control or symptom relief. The data do not support that expectation at ~10% weight loss in older adults with persistent disease. Adjunctive approaches — catheter ablation, antiarrhythmic medications, or potentially GLP-1 receptor agonists (the class of injectable weight-loss drugs including semaglutide, which can achieve weight reductions of 15–20%) — are the appropriate referral pathways when rhythm control is the clinical goal.

For dietitians advising AF patients:

The practical guidance is a shift in framing rather than a reversal of recommendations. Continue supporting weight loss in overweight or obese AF patients: the cardiovascular benefits are real and the safety in older adults is confirmed by this trial. Adjust the patient conversation to be honest about what weight loss can and cannot deliver for the arrhythmia itself — particularly in patients over 65 with long-standing persistent AF who may be expecting weight loss to reduce their AF burden or symptoms. Coupling dietary intervention with timely referral for electrophysiology evaluation is more consistent with the current evidence than positioning diet as a standalone rhythm therapy.

The open question — whether earlier intervention (before AF becomes persistent) or greater weight loss magnitude (≥15%) would produce the AF benefits that LOSE-AF could not — remains unanswered and worth watching in future trials.

Study label: Randomized controlled trial. The JAMA full text was behind a paywall (jamanetwork.com blocked access); this article is based on the structured PubMed abstract, the JAMA editorial accessed via ResearchGate, CardiologyNowNews coverage including direct author quotes, and Medical Dialogues reporting. The complete list of authors beyond the first three and the full limitations section from the paper could not be independently verified from public sources. Conflict-of-interest disclosures for all authors were not confirmed. Trial registration: ClinicalTrials.gov NCT03713775. Funding: NIHR Oxford Biomedical Research Centre and British Heart Foundation.

Cover image: AI-generated illustration — cardiac rhythm printout beside a measured meal portion, depicting the central tension of the LOSE-AF trial.

Weight loss didn't improve AF symptoms in older adults — here's what that means for dietary advice

Who was studied and what was the intervention

The primary result: no change in AF symptoms

Weight loss was real and durable — AF improvement was not

Why LOSE-AF and an earlier trial reached opposite conclusions

Limitations

The dietary takeaway — a concrete recommendation

参考来源