Text messages more than doubled quit rates; fitness texts fell flat

Smoking, inactivity, poor diet and heavy drinking drive the world's biggest killers, the non-communicable diseases, and almost everyone at risk owns a mobile phone. That makes phone-delivered support an enormously tempting public health tool: cheap, scalable and always in a pocket. But temptation is not evidence, and the randomised trials of phone-based behaviour change had not been swept up and weighed together across all four behaviours at this scale. This review set out to establish which phone-delivered interventions actually move the behaviours they target, and which only sound as if they should.

The authors searched MEDLINE, EMBASE, Global Health and CINAHL for randomised controlled trials published between January 1990 and January 2016. They found 71 eligible trials: 18 on smoking cessation, 15 on physical activity, 3 on diet, 25 combining activity and diet, 2 covering activity, diet and smoking together, and 8 on harmful alcohol consumption. Two authors extracted the data, and comparable outcomes were pooled across trials. Only 4 of the 71 trials were judged to be at low risk of bias.

Text message support for quitting smoking stood out: pooled trials showed a relative risk of 2.19 for biochemically verified continuous abstinence and 1.51 for verified seven-day abstinence, with no adverse events reported and no disagreement between the studies. SMS-based diet and activity programmes cut the incidence of diabetes, with a pooled relative risk of 0.67, and improved percentage weight change and triglycerides, though end-point weight itself did not differ significantly. A trial of an SMS physical activity intervention found no difference in peak oxygen uptake at three months, and activity trials reporting outcomes at three months or longer showed no benefit. The alcohol reduction results were inconclusive.

The review draws a sharp line through the mobile health hype: the same delivery channel works powerfully for one behaviour and barely at all for others. For smoking, it confirms text support as one of the better-evidenced digital interventions available, with consistent results across trials. For physical activity, it complicates the assumption that nudges on a screen translate into fitness. The authors note that the most promising interventions still need adequately powered trials measuring illness and death rather than just behaviour, and the January 2016 search cut-off means the modern app era sits outside this evidence.

If quitting smoking is on your list, this is a solid reason to think simple text-based support programmes are worth something: verified quit rates more than doubled across the pooled trials, with no harms reported. If you were hoping a stream of motivational texts would transform your training, the trials here offer no reason for optimism beyond three months. For diet, the honest summary is modest benefit at best, and the biggest outcomes, actual disease and death, remain untested for even the most promising programmes.