Soy Protein, Walking, and Frailty: What the Evidence Shows

Walking and exercise have a better claim on frailty prevention than soy protein does. A 12-week randomized trial found a potentially useful subgroup signal for high-soy-protein soymilk, but the intervention did not significantly improve its prespecified walking-speed outcome across the full group. Exercise-based programs, meanwhile, reduced frailty incidence in a 2024 review of randomized trials.

That distinction matters for anyone trying to preserve independence with age. Protein may support training, especially resistance training. It has not earned a place as a stand-alone frailty treatment.

Why walking speed deserves attention

Frailty is a clinical state of reduced reserve: a minor illness, fall, or period of inactivity can produce a disproportionate loss of function. Gait speed sits near the center of that problem because it captures several systems at once, including leg strength, balance, coordination, cardiovascular capacity, and confidence while moving.

In an eight-year longitudinal study, de Souza and colleagues followed 3,637 older adults initially free of disability in basic activities of daily living and 3,696 free of disability in instrumental activities. Low gait speed identified later disability risk better than frailty and sarcopenia classifications. This does not mean a faster walking test prevents disability. It means gait speed is a meaningful functional marker, not a decorative number on a smartwatch dashboard.

The soy trial: a signal, not a verdict

The direct soy evidence comes from a 12-week randomized controlled trial by Sato et al. in community-dwelling Japanese older adults with pre-frailty or frailty. Participants received continuous high-soy-protein soymilk or a control beverage, and the researchers assessed ordinary walking speed.

The primary result was null: the full analyzed cohort did not show a significant advantage for the soy intervention on the prespecified walking-speed endpoint. The study did report exploratory subgroup benefits among participants who began with an ordinary walking speed of at least 1.0 m/s and among those taking at least 5,000 steps per day.

Those subgroup findings are interesting because they raise a plausible question: perhaps extra protein helps people who already have enough movement capacity to use it. But subgroup analyses after a null main result can also arise by chance. The trial cannot establish that soy protein improves mobility for pre-frail or frail adults generally, and its Japanese community-dwelling population may not translate cleanly to different diets, care settings, or populations.

More steps track with less frailty

The larger walking story is persuasive, though less direct. In a 2021 observational study by Lefferts et al., older adults with hypertension who accumulated more total steps and more aerobic steps had lower frailty prevalence and lower incident frailty. The study used objectively measured step counts, making it more useful than a vague question about whether someone considers themselves active.

Still, this is association, not proof that adding steps will itself prevent frailty. People who are already less frail can walk more, and factors such as illness, pain, socioeconomic circumstances, and baseline fitness can affect both step count and future frailty. The hypertensive study population also limits generalization to healthier adults or people with more severe impairment.

Even with those limits, step count is a sensible tracking metric. It measures a behavior tied to daily function, and the soy trial's subgroup pattern suggests that baseline activity may influence whether a nutrition intervention has any detectable effect.

Exercise beats nutrition alone in prevention trials

The strongest evidence in this group of studies comes from Eidam and colleagues' 2024 systematic review and network meta-analysis of randomized interventions to prevent frailty in adults aged 60 and older who were non-frail or pre-frail. Exercise-based interventions significantly reduced frailty incidence and improved gait speed. Nutritional interventions alone did not significantly prevent frailty onset.

The nutrition-alone estimate illustrates why confidence should stay modest: its relative risk was 1.16, with a 95% confidence interval of 0.33 to 4.10. That interval includes no effect and spans effects in both directions, so it does not support a protective claim. It also signals considerable imprecision.

This does not prove nutrition is irrelevant. It says that nutrition-alone programs have not shown reliable frailty prevention in the trials included in this review. Exercise changes the physical task that frailty threatens: standing up, walking, climbing stairs, recovering balance, and carrying out daily life.

Protein appears more useful alongside training

A 2026 systematic review and meta-analysis by Xie et al. examined nine randomized trials involving 496 older adults with sarcopenia. Resistance training combined with amino acid-based supplementation improved handgrip strength, gait speed, Short Physical Performance Battery scores, and sit-to-stand performance. It did not significantly improve skeletal muscle mass index or appendicular skeletal muscle mass.

The certainty of evidence was low to very low, and the interventions were not soy-specific. Even so, the pattern fits the practical biology: a nutritional input may help most when resistance training provides a reason for the body to adapt. The findings do not justify replacing strength training with a protein shake.

Protein also has plausible biological effects that should not be mistaken for functional proof. A systematic review and meta-analysis of 31 randomized trials, published by Prokopidis et al., found that soy protein supplementation reduced circulating TNF-alpha in older adults. Chronic inflammation contributes to age-related muscle decline, but lower TNF-alpha does not guarantee faster gait, lower disability, or less frailty. A biomarker can move while the outcome people actually care about stays put.

What to do with this evidence

If preserving function is the goal, build the plan around regular walking and structured exercise, with resistance training where medically appropriate. Track habitual steps and periodically measure gait speed under the same conditions. These metrics can reveal declining function early enough to discuss it with a clinician or physical therapist.

• Use higher step counts as a practical activity target and monitoring tool, while recognizing that the available steps study cannot prove causation.
• Prioritize exercise programs because randomized evidence supports their role in reducing frailty incidence and improving gait speed.
• Consider adequate dietary protein, including soy if it suits your diet and health needs, as an adjunct to training rather than a substitute for it.
• Treat high-protein soymilk as unproven for frailty treatment until larger trials replicate a benefit in the full study population.

Protein may help support the work. Walking and resistance training are still the work.

The evidence does not support a soy-first anti-frailty protocol. It supports an exercise-first approach, with nutrition used to reinforce it. The one soy trial offers a reason to study the question further; it does not yet offer a reason to retire your walking shoes.