The test is familiar to anyone who has been worked up for dizziness: lie down for a few minutes, stand up, and let someone take your blood pressure. The diagnosis usually rests on what happens in the first three minutes. A cohort study published on 28 October in BMC Geriatrics argues that the more useful number may be how long the pressure takes to come back — and that the standard definition, applied on its own, missed the people who went on to fall.

The paper comes from the geriatric research institute at Ulm University Medical Center, using the ActiFE cohort, a population-based study of community-dwelling adults in and around Ulm, Germany. The analysis rests on 509 participants examined at the study’s seven-year follow-up in 2016 and 2017. Their median age was 79, 44% were women, and all could walk through a room on their own, with or without a device.

What was measured, and for how long

Blood pressure was taken after ten minutes lying down, then immediately on standing and again at one, two and three minutes. The researchers sorted participants several ways. Orthostatic hypotension in the consensus sense meant a systolic drop of at least 20 mm Hg, a diastolic drop of at least 10 mm Hg, or both. Systolic orthostatic hypotension counted only the systolic criterion. Orthostatic hypertension meant a rise of the same size in either direction. Then each person with a fluctuation was classified by recovery: resolved by the one-minute reading, still present at one minute but gone by three, or never resolved inside the three-minute window.

Falls were not recalled at the end of the year. Participants were given a weekly fall calendar, the study centre collected it every three months and sent reminders every six weeks, and the outcome analysed was whether someone fell at least once over the following twelve months. That prospective design is one of the paper’s strengths: it reduces the recall problem that weakens studies asking older adults to remember how often they fell.

Prevalences were 26.9% for orthostatic hypotension, 16.5% for the systolic-only version, 15.5% for orthostatic hypertension and 9.0% for its systolic-only version. A third of the sample — 168 people — reported at least one fall during follow-up.

The result sat in the recovery minute

Pooled across everyone who met the consensus definition, orthostatic hypotension was not associated with falls: the adjusted odds ratio was 0.96, with a 95% confidence interval from 0.58 to 1.58. The systolic-only definition did clear the bar, at an adjusted odds ratio of 2.00 (1.15 to 3.47), with 45.2% of that group falling versus 18.1% of the reference group.

Recovery time sharpened the picture. Among people whose drop had resolved by one minute, the fall rate was 25.2%; among those still low at one minute and recovered by three, 54.2%; among those who never recovered within three minutes, 60.0%. Adjusted odds ratios were 3.11 (1.21 to 8.01) for delayed recovery and 4.04 (1.07 to 15.26) for no recovery under the consensus definition, and 4.73 (1.71 to 13.10) and roughly fivefold under the systolic-only definition.

The authors are explicit that these subgroups are thin. Delayed recovery under the systolic definition held 19 people, 12 of whom fell; the no-recovery group held eight, five of whom fell. An odds ratio built on eight people carries a confidence interval wide enough to span almost anything, and the paper says so: the recovery subgroups “and wide confidence intervals require cautious interpretation and confirmation in larger studies.”

Orthostatic hypertension went the other way — nothing. Its adjusted odds ratio for falls was 1.05 (0.58 to 1.92), the systolic-only version 1.02 (0.48 to 2.16), and breaking it down by recovery time did not produce a signal either. The authors note that this is the first study to look at recovery patterns in that group, and that a sustained rise on standing appears to behave differently from a sustained fall.

Why the systolic number drew the attention

Part of the argument is mechanical. Standing shifts blood into the legs; the body answers by tightening vessels and lifting heart rate to keep stroke volume and cardiac output up, which is what maintains perfusion to the brain. A systolic decline may be the better marker of that compensation failing, since diastolic pressure tends to stay flat or creep up during standing rather than fall. The paper points out that isolated diastolic orthostatic hypotension is rare and that some researchers have proposed dropping the diastolic criterion altogether.

It also fits a wider pattern in the literature the authors cite: earlier studies have found the consensus definition inconsistently predictive of falls, while systolic-only definitions and delayed recovery have done better. None of that makes recovery time a diagnostic test. It makes it a hypothesis worth testing properly, which is what the paper’s final paragraph asks for.

What this study cannot tell you

The sample is community-dwelling and relatively high-functioning; the authors warn their findings should not be extended to people in residential care, or to those with mobility or sensory limitations. Symptoms and heart-rate responses could not be analysed because too much of that data was missing — so the study says nothing about whether the people with delayed recovery felt dizzy. Blood pressure was taken with an intermittent cuff rather than beat-to-beat monitoring, so swings between measurements were invisible. The test moved people from lying to standing, while most everyday transitions are from sitting. And the fall calendar recorded that a fall happened, not where or when, which means the authors cannot tie any individual fall to a blood-pressure drop.

There is also no trial here, and none anywhere that the paper cites, showing that intervening on recovery time prevents falls. That question remains open.

What the guidance already says

The scale of the problem is not in dispute. The Centers for Disease Control and Prevention reports that more than one in four adults aged 65 and older falls each year, that fewer than half tell a doctor, and that falling once doubles the chance of falling again. Its figures put older-adult falls at about 4.5 million emergency department visits a year, 1.4 million of them ending in hospitalisation, with roughly 319,000 hip-fracture hospitalisations annually.

The CDC’s prevention advice is unglamorous and does not depend on this paper: ask a clinician to assess fall risk; have a doctor or pharmacist review medicines, including over-the-counter ones, for anything that can cause dizziness or drowsiness; do strength and balance work, with tai chi named as an example; get vision checked yearly; and reduce trip hazards, add grab bars and railings, and improve lighting at home. Several of the drugs the Ulm analysis adjusted for are exactly the sort that show up on a medication review.

MedlinePlus sets the consumer frame: a reading of 90/60 or lower counts as low blood pressure, some people run low with no symptoms at all, and it becomes a problem when it causes dizziness, fainting or, in extreme cases, shock. Its summary notes that symptoms often appear on standing up too quickly.

Put those together and the practical translation of this study is modest. If you or someone you care for gets light-headed on standing, or has already fallen, that is a conversation for a clinician — and the medication list is the thing to bring to it. Whether the blood pressure is measured at one minute or three is a decision for whoever is doing the measuring, not a self-test to run at home. As with everything on this site, this is reporting on a study, not medical advice; anyone with symptoms or a fall should talk to a healthcare professional rather than adjust anything on their own.

For readers interested in the other end of the same question — how much walking intensity matters in older age — our earlier piece on a frailty trial of brisk versus casual walking covers the trial evidence.