Bariatric surgery, compared to medical treatment, reduces morbidity at all ages but does not reduce mortality in patients aged < 43 years, especially if diabetes mellitus is present: a post hoc analysis of two retrospective cohort studies

Background and aims: Bariatric surgery (BS) reduces long-term mortality in comparison with medical treatment of obesity. Some studies indicate that this effect is significant for patients above mean age in different cohorts, but not for younger patients. These findings raise the question whether morbid obese patients should undergo BS as soon as possible, or whether patients might undergo surgery later in their life. Methods: We performed a post hoc analysis of two studies; we evaluated surgery-related long-term mortality in: (1) the whole cohort [857 surgery patients (163 diabetes) vs. 2086 controls (512 diabetes)]; (2) patients above mean age [&gt; 43 years, 427 surgery patients (133 diabetes) vs. 1054 controls (392 diabetes)]; (3) patients below mean age [≤ 43 years, 432 surgery patients (30 diabetes) vs. 1032 controls (120 diabetes]. Then, we analyzed age-related long-term mortality in the whole cohort, as well as in surgery patients and in controls. Finally, we analyzed incident diseases (diabetes, cardiovascular disease, and cancer) as a function of surgery versus no-surgery and of mean age. Results: Surgery patients, compared with controls receiving standard medical/dietary treatment, had reduced mortality in the whole cohort (HR = 0.45, 95% CI 0.33–0.62, p = 0.001) and in the study group aged &gt; 43 years (HR = 0.39, 95% CI 0.28–0.56, p = 0.001), but not in the study group aged ≤ 43 years (HR = 0.87, 95% CI 0.42–1.80, p = 0.711). Reduced mortality was observed in non-diabetic and diabetic patients aged &gt; 43 years (HR = 0.37, 95% CI 0.23–0.62, p = 0.001 and HR = 0.45, 95% CI 0.27–0.74, p = 0.002, respectively) who underwent bariatric surgery. In contrast, in patients aged ≤ 43 years, no significant protective effect of bariatric surgery appeared in non-diabetic patients (HR = 0.64, 95% CI 0.24–1.71, p = 0.371), and mortality increased, almost significantly, in diabetic patients aged &lt; 43 years (HR = 2.87, 95% CI 0.96–8.56, p = 0.058), and even more in diabetic patients aged 33–43 years; HR = 4.99, 95% CI 1.18–21.09, p = 0.029). As expected, age-related mortality was increased in the whole cohort (HR = 7.23, 95% CI 5.14–10.17, p = 0.001), in non-diabetic and diabetic controls (HR = 8.55, 95% CI 5.77–12.68, p = 0.001, and HR = 3.76, 95% CI 1.97–7.18, p = 0.001, respectively). The effect of aging was slightly reduced in surgery patients (HR = 3.76, 95% CI 1.87–7.58, p = 0.001), while it was not significant in diabetic surgery patients (HR = 0.70, 95% CI 0.26–1.90, p = 0.88), further emphasizing that diabetes per se has a strong negative effect on survival, also with concomitant bariatric surgery. In a supplementary analysis, HRs did not change when surgery and control parents were matched for the presence of diabetes. Incident diseases (cardiovascular, diabetes, and cancer) were less frequent in surgery than in control patients, irrespective of age. Conclusion: Bariatric surgery reduces long-term mortality in comparison with medical treatment when performed in patients aged &gt; 43 years, but not in younger patients, where it is neutral or could even increase mortality; reduction in morbidity occurs at any age.