Playing Football With a Chronic Disease? A Major Meta-Analysis Weighs the Evidence

Could a kickabout twice a week be a genuine prescription for people living with hypertension, diabetes, or cancer? A new systematic review and meta-analysis published in Frontiers in Public Health suggests that recreational football — small-sided, non-competitive, group-based play — delivers measurable physiological benefits across a range of chronic disease populations, while also sounding a careful note of caution about how strong the evidence really is. The review, led by Kun Wei of Henan Sport University and Gang Chen of Nanjing University, pooled controlled trials of football-based exercise in patients with chronic conditions and assessed both the health effects and the practical feasibility of the format.

The research team searched PubMed, Web of Science, and the Cochrane Library from their inception to 6 February 2026, with additional searches of Scopus and Embase conducted on 17 August 2026. After screening more than 1,300 records, they identified 22 study reports representing 17 unique trials — 15 randomised and 2 non-randomised controlled trials — encompassing 588 unique participants. The clinical populations spanned hypertension, prediabetes, type 1 and type 2 diabetes, breast cancer, prostate cancer, and broader metabolic risk profiles. Participants ranged in age from roughly 14 to 67 years, and interventions ran from 8 weeks to 12 months, typically delivered two to three times per week at moderate-to-vigorous intensity, with net playing time between 7 and 45 minutes per session.

Methodologically, the analysis was more sophisticated than most in this field. Because some trials contributed multiple effect estimates — through multiple intervention arms, linked reports from the same cohort, or repeated measures of the same outcome — the authors used three-level random-effects meta-analyses that treated effect sizes as nested within trials. They then applied cluster-robust (CR2) variance estimation, clustering on the independent trial, to every outcome as a robustness check, and drew conclusions only where the two inferential procedures agreed. Certainty of evidence was rated with the GRADE framework, and prediction intervals were reported to show how variable the true effects might be across different settings.

The headline findings are striking for a group of patients often excluded from vigorous team sport. Recreational football was associated with significant improvements in cardiorespiratory fitness, with a standardised mean difference (SMD) of 0.70 for peak oxygen uptake (VO₂peak) and an even larger 1.43 for Yo-Yo Intermittent Endurance Test performance. Cardiovascular outcomes also improved: diastolic blood pressure fell (SMD −0.53), mean arterial pressure dropped (SMD −0.89), and resting heart rate showed the largest effect in the entire synthesis (SMD −1.01). In clinically interpretable units, the pooled mean differences amounted to roughly a 5 mmHg reduction in mean arterial pressure and a 5.7 beats-per-minute reduction in resting heart rate.

Body composition and blood lipids moved in the right direction too. Body fat percentage declined significantly (SMD −0.36), muscle mass increased modestly (SMD 0.22), and both triglycerides (SMD −0.56) and total cholesterol (SMD −0.69) fell. Notably, body weight, BMI, and waist circumference did not change significantly — a pattern the authors attribute to the insensitivity of scale-based measures to compositional change when fat is lost and lean tissue is gained, particularly in patients taking medications with weight-active effects. One prostate cancer trial, for example, reported a between-group lean body mass difference of about 0.7 kilograms after 12 weeks of football training in men undergoing androgen deprivation therapy.

The bone-health results tell a subtler story. Plasma osteocalcin, a marker of bone formation, rose significantly (SMD 1.07), and P1NP, another formation marker, increased under the primary analysis (SMD 1.28). Yet DXA-derived bone mineral density and content showed no significant changes, and the P1NP effect did not survive cluster-robust inference. The authors interpret this as evidence of early anabolic signalling in response to the mechanical loading of football — the jumps, turns, accelerations, and decelerations that small-sided games demand — without enough training exposure to produce measurable structural remodelling in older, clinically vulnerable cohorts. Bone turnover markers can respond within weeks, whereas density changes typically require longer and are most visible at mechanically loaded lower-limb sites.

Several seemingly positive findings proved fragile under scrutiny. Borderline effects on fat mass, systolic blood pressure, and LDL cholesterol were not retained under cluster-robust inference, and effects on body fat percentage, fat mass, muscle mass, systolic pressure, and LDL were sensitive to the assumed pre–post correlation used to derive change-score statistics. No significant pooled effects emerged for HbA1c, fasting glucose, fasting insulin, HDL cholesterol, or C-reactive protein. Across 27 outcomes, the statistical interpretation was unchanged for 22 under the robust procedure — but certainty of evidence was rated low to very low for most outcomes, and prediction intervals for several significant findings crossed zero, meaning the average benefit was not reproduced consistently across diagnoses and programmes.

On feasibility, the picture is encouraging but not uniform. Mean attendance across 13 independent trials was 82.7 percent, and mean attrition across 14 trials was 15.6 percent — respectable figures for exercise interventions in clinical populations, and consistent with the idea that the social structure, team affiliation, and sheer enjoyment of football help sustain engagement where solitary treadmill work often fails. Adverse events were summarised narratively because quantitative pooling was impossible: most reported events were musculoskeletal, including muscle soreness, strains, joint pain, and lower-limb injuries, with occasional serious injuries such as fibular fractures and Achilles tendon ruptures. Several trials noted that injuries clustered in the early phase of training, underscoring the importance of progressive exposure, warm-up, and qualified supervision.

The authors are explicit that these results should not be read as evidence that football can replace established exercise prescriptions or as a disease-specific therapy. The included populations differ enormously in pathophysiology, treatment burden, medication status, and baseline fitness, and the number of studies within any single diagnostic category was too small for firm conclusions. In hypertension, blood-pressure effects were inconsistent, particularly among patients already on antihypertensive medication, where residual headroom for further lowering may be limited. In diabetes, glycaemic outcomes were largely unchanged at the pooled level, plausibly because fasting indices and HbA1c are relatively insensitive to modest exercise effects in pharmacologically controlled patients. Cancer-related evidence came from small, highly selected samples with variable attendance.

Nevertheless, the synthesis marks a milestone for the ‘Football is Medicine’ platform, which has argued for years that small-sided games deliver a uniquely multimodal stimulus — cardiovascular, neuromuscular, and osteogenic load within a single session — that continuous aerobic training and isolated resistance exercise cannot match. The physiological profile is coherent: heart rates of 80 to 90 percent of maximum, repeated high-intensity efforts, and weight-bearing impacts together explain the concurrent gains in fitness, heart rate, body fat, and bone formation markers. For appropriately screened patients who prefer group-based, game-based activity and can participate safely under supervision, the authors conclude that recreational football may be considered a supervised adjunctive exercise option — particularly where cardiorespiratory fitness is the target. Adequately powered, diagnosis-specific trials with standardised reporting of training dose, adherence, and adverse events are the next step before football earns a formal place in the clinical exercise toolkit.

Subject of Research: Effects and feasibility of recreational football exercise in patients with chronic diseases

Article Title: Recreational football as medicine? A meta-analysis of its effects and feasibility in patients with chronic diseases

Article References: Wei, K., Chen, G., Wang, T., & Wang, B. (2026). Recreational football as medicine? A meta-analysis of its effects and feasibility in patients with chronic diseases. Frontiers in Public Health, 14, Article 1869982. https://doi.org/10.3389/fpubh.2026.1869982

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1869982

Keywords: recreational football, meta-analysis, chronic disease, cardiorespiratory fitness, blood pressure, hypertension, type 2 diabetes, cancer survivors, bone health, exercise as medicine, body composition, feasibility