Simple Grip Test, Adjusted for Body Size, Predicts Falls in Older Adults

A simple squeeze of the hand may reveal far more about an older person’s risk of falling than doctors previously realized, but only if the measurement is interpreted the right way. A new prospective cohort study drawing on data from the China Health and Retirement Longitudinal Study, known as CHARLS, suggests that handgrip strength becomes a markedly better predictor of future falls when it is adjusted for body mass index. The research, published in BMC Geriatrics by Lixia Liu, Zhijie Li, Qian Dai and Qinqin Wu, followed nearly five thousand older Chinese adults and found that the ratio of handgrip strength to body mass index, abbreviated as HGS/BMI, consistently flagged elevated fall risk in both men and women, while the conventional absolute grip measurement failed to do so in women altogether. The finding carries immediate practical weight for clinicians who rely on grip dynamometry as a quick, inexpensive window into whole-body muscle function.

Falls are among the most consequential health events in later life. They are a leading cause of injury-related disability and death in older adults, and they trigger cascades of lost independence, hospitalization and rising care costs. Because falls are so common and so damaging, geriatric researchers have long searched for screening tools that can identify high-risk individuals before the first fracture or head injury occurs. Handgrip strength has been a centerpiece of this effort. Measured with a handheld dynamometer, it takes seconds, requires no specialized facility, and correlates with overall skeletal muscle strength better than almost any other single bedside test. International consensus frameworks for sarcopenia, the age-related loss of muscle mass and function, have accordingly adopted grip strength as a core diagnostic criterion, and the Asian Working Group for Sarcopenia, whose 2025 update the new study explicitly references, has been shifting emphasis from muscle mass toward muscle function.

Yet absolute grip strength has a stubborn statistical blind spot: body size. A taller, heavier person generally produces more absolute force than a smaller one, simply because larger bodies come with larger muscles. When a clinician compares a raw grip reading against a population cutoff, that reading conflates true muscular quality with sheer body dimensions. This is the same conceptual problem that plagued body weight measurements before body mass index normalized weight for height. The research team behind the new analysis reasoned that dividing grip strength by body mass index, expressed in kilograms of force per unit of BMI, would strip out the size confound and yield a cleaner index of relative muscle function, one that might track functional outcomes such as falls more faithfully than the raw number.

To test that hypothesis, the investigators turned to CHARLS, a nationally representative longitudinal survey of middle-aged and older Chinese residents. They anchored their analysis at the 2011 baseline wave, when muscle strength was assessed in thousands of participants, and then tracked self-reported falls during subsequent follow-up. Their analytic sample comprised 4,978 adults aged over the study’s older-adult range, split almost evenly between 2,510 men and 2,468 women. Baseline demographic variables and physician-confirmed chronic diseases were included as covariates in multivariable logistic regression models, allowing the team to isolate the independent contribution of each strength measure. During follow-up, 14.80 percent of the men and 22.37 percent of the women reported falling, a sex gap consistent with the international literature showing that older women fall, and are injured by falls, more often than older men.

The results split sharply along the line dividing absolute from adjusted strength. When handgrip strength was entered into the models as a continuous variable, each one-kilogram increase in absolute grip was associated with a 2.3 percent reduction in fall risk among men, with an odds ratio of 0.977 and a 95 percent confidence interval of 0.963 to 0.992, a statistically significant effect at P equals 0.003. Among women, however, absolute grip strength showed no significant association at all, with an odds ratio of 0.991 and a confidence interval spanning unity, from 0.975 to 1.008, at P equals 0.297. In other words, for the women in this cohort, the raw dynamometer reading carried essentially no predictive information about future falls once other factors were accounted for.

The BMI-adjusted ratio told a different and far more consistent story. Higher continuous HGS/BMI values were significantly associated with lower fall risk in both sexes, with nearly identical effect sizes: an odds ratio of 0.698 per unit increase in men, with a 95 percent confidence interval of 0.513 to 0.951 at P equals 0.023, and an odds ratio of 0.700 in women, with a confidence interval of 0.504 to 0.973 at P equals 0.034. The symmetry of these estimates is striking. Where absolute grip strength produced a significant result in one sex and a null result in the other, the adjusted index produced essentially the same protective association in both, suggesting that normalizing for body size removed a source of sex-specific distortion rather than introducing one.

Categorical analyses reinforced the pattern. When participants were grouped by conventional low versus high grip strength, men in the low-grip category showed a significantly elevated fall risk, with an odds ratio of 1.788 and a confidence interval of 1.354 to 2.360 at P below 0.001, but the corresponding association in women faded toward non-significance, with an odds ratio of 1.246 and a confidence interval of 0.966 to 1.608 at P equals 0.091. The decisive test came with quartile-based stratification of the HGS/BMI ratio. In that analysis, adults in the lowest quartile of BMI-adjusted grip strength remained independently and significantly more likely to fall than those in the highest quartile, in both men, with an odds ratio of 1.391 and a confidence interval of 1.015 to 1.905 at P equals 0.040, and women, with an odds ratio of 1.331 and a confidence interval of 1.006 to 1.762 at P equals 0.045. Across continuous and categorical modeling strategies alike, the adjusted index retained its predictive value where the absolute measure did not.

The mechanistic logic behind this divergence is worth unpacking. Falls are not caused by weak hands; they are caused by failures of the neuromuscular system to recover balance, by slow reaction times, by frailty and by the inability of leg and trunk muscles to generate force quickly enough. Grip strength works as a screening proxy precisely because it samples the same underlying physiology of age-related muscle decline. But when that proxy is expressed in absolute terms, a heavier older adult can post a respectable raw grip score while carrying a disproportionate burden of fat mass relative to functional muscle, a body-composition pattern sometimes described as sarcopenic obesity. That individual’s true functional reserve is worse than the raw number suggests, and the unadjusted measurement masks it. Dividing by BMI partially corrects for this, penalizing strength that is merely the product of a large frame and rewarding strength that is high relative to body size, which is closer to what balance recovery actually demands.

The clinical implications follow directly. Screening programs that use absolute grip cutoffs may systematically underestimate fall vulnerability in women and in smaller or heavier adults, precisely the groups in whom fall-related injuries are most devastating. The authors argue that HGS/BMI offers superior clinical utility and describe it as a rational conceptual extension of the evolving AWGS 2025 framework, which has been moving from muscle mass toward muscle function as the operative concept in sarcopenia care. Because the adjustment requires nothing more than a scale, a stadiometer and the dynamometer already in use, adopting the ratio imposes no additional cost or burden on busy clinics or community screening programs. A nurse could compute it on the spot.

Certain caveats temper the enthusiasm. The outcome was self-reported falling, which is vulnerable to recall error, and the cohort is Chinese, so cutoffs and associations may need recalibration in other populations and health systems. As with any observational cohort, residual confounding by unmeasured health conditions or activity levels cannot be excluded, and the published version was shared early as a citable accepted manuscript subject to further editorial refinement. Even so, the central message is robust and actionable: the same dynamometer reading becomes a better crystal ball when it is divided by body size. For a world with rapidly aging populations and enormous fall-related burdens, a free arithmetic step that rescues the predictive power of a thirty-second bedside test is the kind of quiet methodological upgrade that can ripple through screening guidelines, and through the lives of the older adults those guidelines protect.

Subject of Research: BMI-adjusted handgrip strength as a predictor of fall risk in older adults

Article Title: BMI-adjusted handgrip strength as a functional index to assess fall risk among older adults

Article References: Liu, L., Li, Z., Dai, Q., & Wu, Q. (2026). BMI-adjusted handgrip strength as a functional index to assess fall risk among older adults. BMC Geriatrics. https://doi.org/10.1186/s12877-026-08394-7

Image Credits: AI Generated

DOI: 10.1186/s12877-026-08394-7

Keywords: handgrip strength, body mass index, falls, older adults, sarcopenia, relative muscle strength, CHARLS, prospective cohort study, geriatrics, fall risk screening, muscle function, BMC Geriatrics