- Design
- retrospective longitudinal analysis with linear mixed-effects modelling, eye as random effect, 2012 to 2023
- Population
- 2,655 measurements from 154 eyes of 77 patients with stable treated primary open-angle glaucoma; mean age 65.8 years
- Primary outcome
- intraocular pressure by Goldmann applanation
- Effect
- ambient temperature estimate -0.029 (P = 0.003); mean blood pressure 0.021 (P < 0.001)
Eleven years of Goldmann applanation readings from a Kyoto clinic were analysed for the effect of weather on intraocular pressure in stable treated primary open-angle glaucoma. Inclusion required no ocular surgery, no medication change, age 40 or over, and at least three visits a year for three years — a design that isolates environmental variation by removing the clinical variation. A linear mixed-effects model with eye as the random effect adjusted for ocular biometry, systemic variables, five weather measures, season, time of day and drop count.
Ambient temperature was negatively associated with pressure (estimate -0.029, P = 0.003) and mean blood pressure positively (0.021, P < 0.001). Number of drops, glaucoma type and time of day also had significant effects, with season showing a trend.
The coefficients are small: about 0.29 mmHg for a 10-degree temperature difference, and roughly 0.2 mmHg for a 10 mmHg change in mean blood pressure. Neither is going to change a treatment decision on its own. What they do is explain part of the variation that gets attributed to treatment or to progression — a patient reviewed in January and again in May has had a temperature change that moves the number slightly, independent of anything that happened to their disease.
That has a specific use in India, where the seasonal temperature range in most of the country is far wider than in Kyoto, and where the clinic is often air-conditioned while the waiting room is not. The honest conclusion is not to correct pressures for the weather but to be slower to act on a small change between visits taken in different seasons, and to keep measurement conditions as constant as the clinic allows.
- Be slower to act on a small pressure change between visits taken in different seasons
- Record the time of day and keep review appointments at a consistent time where possible
- Note the patient's blood pressure alongside the pressure reading; the two moved together here
- Do not apply a temperature correction to readings — the effect is too small and too imprecisely estimated for that
- Base treatment escalation on a trend across several visits rather than on one reading
Why it matters
Part of what looks like a response to treatment, or a failure of it, is the month the reading was taken in.
Don't overread it
The effect is a fraction of a millimetre of mercury and was observed in one temperate city — it does not justify correcting measurements.
The statistics, in plain English
An estimate of -0.029 means about 0.03 mmHg of pressure per degree of temperature, so roughly 0.3 mmHg across a 10-degree swing — statistically detectable across 2,655 measurements, and far below what a clinician would act on. The mixed-effects model treats each eye as its own baseline, which is the right way to handle repeated measurements from the same patients. This is a single Japanese clinic with a temperate range, and the relationship at temperatures common in much of India was not observed.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for ophthalmology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free