Diabetes is spreading fastest worldwide in low- and middle-income countries, where cash transfers often serve as the backbone of social protection. However, the link between such payments and diabetes risk has been almost unstudied. This work stands apart: nearby in the archive there are studies on glucose monitoring and health forecasting, but none treats cash assistance as an intervention. This is the first attempt to connect social policy with glucose biomarkers at the population level.
The researchers used data from the HAALSA Indepth cohort in the rural district of Agincourt, South Africa. They reconstructed the month-by-month history of eligibility for the two largest cash transfers, the old-age pension and the child grant, for each household since 2000. Then they calculated the average monthly amount per person, adjusted for inflation. Glucose levels were measured with glucometers in 2014/15 and 2021/22. To analyze the link between payments and the development of diabetes or prediabetes, they used regression models accounting for age, sex, and participant attrition.
Over seven years, diabetes developed in 16.8% of participants who did not have diabetes at baseline, and prediabetes in 21.7% of those initially healthy. Every additional 100 rand (about 6 dollars) per month reduced the risk of prediabetes by 6% (confidence interval 0.90–0.98). For diabetes, there was no association. Quantile regression showed that in people with the highest glucose levels, the reduction in glucose was three times greater than the group average. The results did not depend on the choice of prediabetes criteria or the method of calculating payments.
The work shows for the first time that cash transfers can prevent the early stages of impaired glucose metabolism, but do not halt diabetes that has already begun. This is consistent with the idea of prediabetes as a more plastic state. Unlike neighboring studies that focus on diagnosis or monitoring, here the intervention targets the social determinants of health. The result underscores that fighting diabetes in poor communities requires not only medicine but also economic support.
The topic could evolve toward identifying specific mechanisms: what exactly do the payments change, diet, physical activity, stress, or doctor visits? Randomized trials with different payment amounts and conditions are possible. Studying the threshold beyond which the effect extends to diabetes is promising. It is also important to test whether such programs work in other countries with a similar structure of social assistance.
The results could influence the design of social protection programs in low- and middle-income countries, where diabetes is growing fastest. They also provide an argument for including cash transfers in strategies for preventing noncommunicable diseases.
The next step is to find out which specific lifestyle changes mediate the reduction in prediabetes risk. Then a randomized study with larger payments should be conducted to test whether diabetes can also be affected.
The work addresses a long-standing problem: why socioeconomic factors so strongly influence diabetes risk, and whether this effect can be reversed. It also raises the question of the intervention threshold: at what stage of impaired glucose metabolism is prevention still possible, and when are medications already needed.
The highest glucose level recorded in the study was 30.6 mmol/L, which is three times the diabetes threshold and life-threatening. Yet the average payment per person was only about 21 dollars a month, less than the cost of a single doctor's visit in developed countries.
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.