The study that started present interest in metformin as a potential approach to modestly slow aging was problematic in a number of ways, as examined in a lengthy series of posts at the SENS Research Foundation. It appeared to show that type 2 diabetics on metformin enjoyed a small survival advantage over non-diabetics not taking metformin. Here, researchers look at survival over twenty years, and find an apparent short-term gain in life expectancy over non-diabetics that vanishes after a few years. The hypothesis is that the effects of type 2 diabetes overwhelm the benefits of metformin given time. It remains unclear as to whether metformin can have any meaningful small benefit for non-diabetics; finding out is the goal of the proposed TAME trial.
We examined longevity in type 2 diabetes (T2D) patients treated with metformin therapy and compared them to matched controls and T2D patients treated with sulphonylurea therapy. Looking at individuals over a period of up to twenty years we showed that T2D patients had shorter survival times after first treatment than matched controls. When the study period was artificially truncated, we found a statistically significant benefit of metformin therapy for longevity over matched non-diabetic controls within the first three years. However, this benefit disappeared when we looked over longer periods of time (after five years).
This suggests that benefits of metformin may be short-term only and/or the longer-term benefits of metformin are negated by the life-shortening effects of T2D and associated comorbidities. An alternative explanation is that T2D patients experience better short-term survival outcomes following treatment due to lifestyle adjustment, as recommended by doctors. However, we did not see a benefit to longevity in the short-term for sulphonylurea therapy patients who would presumably be motivated to improve their lifestyle in the same way.
Metformin has been linked to lower mortality due to cancer, and to reduced cardiovascular disease (CVD) risk. Compared to the sulphonylurea therapy group, we did see significantly lower lifetime prevalence of cancer, and lower rates of cardiovascular disease. Excluding individuals with history of cancer and CVD prior to first treatment, these differences were even larger. This finding is supportive of the protective effects of metformin for cancer and CVD compared to other diabetes treatments. However, as we used non-diabetic controls who were matched on cancer and CVD status to the diabetic cases, we are unable to distinguish if there is a benefit of this treatment.