Texts nudged blood pressure down in pooled trials

Cardiovascular disease is the leading cause of death worldwide, claiming 17.9 million lives annually, and the biggest contributors are behavioural: poor diet, inactivity, tobacco and excessive alcohol. Changing those behaviours at scale is difficult and expensive through face-to-face care alone. Text messaging is about as low-tech and low-cost as digital health gets, works on any phone, and requires no app download or wearable. The open question is whether something that simple actually moves clinical risk factors, or merely feels like an intervention.

The review followed PRISMA 2020 guidelines, with searches run on 12 February 2024 across PubMed, MEDLINE, Cochrane, Scopus, Web of Science, Embase and CINAHL using both subject headings and free-text terms for cardiovascular disease and text messaging. From 5,748 identified articles, 22 studies met the inclusion criteria and were carried into the meta-analysis. Only randomised clinical trials were pooled. Outcomes analysed included medication adherence, systolic and diastolic blood pressure, BMI, LDL, HDL, total cholesterol and HbA1c, with pooled effects expressed as mean differences with 95% confidence intervals and heterogeneity reported as I².

Medication adherence improved significantly, with a pooled mean difference of 0.62 and no detectable heterogeneity between trials. Diastolic blood pressure fell by 2.66 mmHg and systolic by 6.12 mmHg, both statistically significant, though heterogeneity was high at 85% and 96% respectively. Nothing else moved: BMI, LDL, HDL, total cholesterol and HbA1c all showed no significant improvement. The pattern suggests the messages worked mainly through behaviours that a reminder can plausibly influence, rather than through broad metabolic change.

This supports the wider case for simple digital nudges as an adjunct to cardiovascular prevention while sharply limiting the claim: adherence and blood pressure yes, lipids and glycaemic control no. The very high heterogeneity for blood pressure means the trials differed substantially in who they recruited, what the messages said and how long they ran, so a single pooled millimetre figure smooths over a lot of variation. It leaves open which message content, frequency and personalisation actually drive the effect, and whether benefits persist once the novelty of the reminders wears off. Long-term follow-up and trials in more diverse populations are the obvious next steps.

If reminders help you stick with something you have already decided to do, this pooled evidence suggests that effect is real and can register in blood pressure readings. It is also a reminder of the ceiling: a text can prompt a behaviour, but it did not shift body weight, cholesterol or blood sugar markers in these trials. That distinction is useful when judging any digital health tool, whether it is a messaging service or an app promising broad metabolic transformation. Blood pressure itself remains one of the more responsive and more consequential numbers to keep an eye on.