Editorial NexoScience
Does worrying about falling keep you upright, or does it warn that a fall is already on the way? Pooled evidence from decades of prospective research now leans hard toward the second answer. It also comes with an awkward extra finding: the thing clinics often reach for, how confident an older person says they feel on their feet, predicted nothing at all.
The warning sign is the worry itself. Researchers pooled 53 prospective studies covering 75,076 adults aged 60 and over, a group roughly the size of a mid-sized city, and in every one of them the question about falling came first and the falls were counted afterwards. Where a study used a blunt single question, usually “Are you afraid of falling?”, people who answered yes had 60% higher odds of falling during follow-up than those who answered no. Where a study used the Falls Efficacy Scale International, a questionnaire that asks how concerned you are about falling while doing ordinary things like getting dressed or walking on a slippery floor, the odds crept up with every extra point, and someone scoring 40 points higher on the full scale carried more than double the odds of falling of someone at the bottom of it.
The findings were published in 2025 in the journal Age and Ageing. The review was registered in advance in the PROSPERO database, which locks in the research question before the data are seen, and it was built to answer one narrow question: does concern about falling predict future falls independently of the risk factors doctors already know about, such as age, sex, physical function and previous falls?
The team only accepted studies that watched people forward in time. Four research databases were searched up to March 2024, and cross-sectional snapshots, trials and case reports were thrown out. What survived had to follow participants for at least six months, with falls logged either as they happened, through diaries or regular phone calls, or recalled at the end. The results were combined as odds ratios, a number that says how much more likely an outcome is in one group than another, and wherever authors had adjusted for other risk factors, it was the adjusted figure that went into the pool.
Balance confidence failed the same test that worry passed. Nine studies measured it with the Activities-Specific Balance Confidence scale, which asks people how sure they are that they won’t lose their footing during everyday tasks. Pooled together, the score showed no relationship with who fell later (odds ratio 0.97, with a range from 0.93 to 1.01 that comfortably includes no effect at all). The researchers were blunt about the practical consequence, writing that balance confidence should not be part of standardised fall risk assessments, and noting that people with low confidence but no active concerns are not necessarily at higher risk.
The likeliest explanation is a loop that feeds itself. Concerns about falling affect up to half of people over 60, and the authors describe the relationship as bidirectional: a fall breeds worry, and the worry then breeds more falls. The mechanism is depressingly ordinary. Worry leads to skipped walks and avoided stairs, and balance behaves like a language, in that the less you speak it the harder it becomes to speak at all. Laboratory work cited in the review points the same way, with anxious walkers moving in an overly cautious style that is not actually safer.
None of this amounts to proof of cause. Every study in the pool was observational, so a worried person who falls may simply be a person whose body is already failing and who has correctly noticed it. The authors leaned on the Bradford Hill criteria, a checklist epidemiologists use to argue that an association is likely causal, but a checklist is an argument, not an experiment, and they rated their own certainty as moderate rather than high. Half the included studies were graded poor quality, most of them for not controlling for obvious confounders or for asking people to remember falls months later, and most excluded anyone with cognitive or neurological impairment, which leaves out a large slice of the people who actually fall. The null result for balance confidence deserves its own asterisk: it rests on 1,728 participants, against nearly 44,000 in the analysis of the single-question measure, and it was downgraded for imprecision.
The open question is whether treating the worry moves the falls. The review could not pool anything on repeat falls or injurious falls, the outcomes that put people in hospital, because the studies defined them too differently to combine. Long-term care residents are almost absent from the evidence, with only four studies recruiting from residential settings. What the authors do recommend is cheap: the short seven-item version of the questionnaire takes under three minutes, and it is already validated. The harder trial has not been run yet, and it would test whether exercise, cognitive behavioural therapy and occupational therapy aimed squarely at the fear can bring the fall rate down with it.
Source: “Does concern about falling predict future falls in older adults? A systematic review and meta-analysis”, published in the journal Age and Ageing (2025). PROSPERO registration CRD42023387212. DOI: 10.1093/ageing/afaf089.
