Key Points:
- There is uncertainty determining who is likely to develop dysglycemia and type 2 diabetes mellitus (T2DM) due to weight gain and adiposity: 80% of those with obesity do not develop T2DM, and 50% of persons with T2DM do not have obesity.
- Targeting weight loss to certain individuals could personalize the prevention of T2DM, helping determine who is most likely to benefit from weight loss therapy to prevent T2DM.
- A total of 445,765 participants in the UK Biobank were analyzed, of whom 28,563 developed T2DM, in order to infer if more T2DM cases are inherited or acquired, to determine if there is optimal timing of T2DM prevention, and to ascertain if certain groups are more likely to benefit from T2DM prevention.
- Given substantial variation in the effect of BMI on A1c depending on multiple variables, individuals may have a “personal weight threshold” for developing T2DM that could be individually targeted.
There is a current need to address the twin epidemics of T2DM and obesity. Obesity increases the risk of T2DM, and T2DM is associated with high morbidity, mortality, and healthcare costs. However, though weight gain and increased adipose tissue increase the risk of developing T2DM, it is difficult to determine exactly who is most susceptible to this disease development. Notably, 80% of those with obesity do not develop T2DM, and 50% of persons with T2DM do not have obesity. By analyzing a large genetic database, the study sought to investigate whether targeting weight loss to certain patients could personalize the prevention of T2DM.
This study analyzed 445,765 participants enrolled in the UK Biobank, of whom 28,563 developed T2DM. For external validation, the study population was compared to 212,351 participants enrolled in FinnGen Biobank study, of whom 29,166 developed T2DM. They first compared the effect of increased BMI on T2DM risk with effect of a 2,037,596 variant polygenic risk score (PGS) for T2DM to infer whether most T2DM cases are acquired or inherited. Next, they compared the effect of random allocation to lifelong exposure to a 1-unit increment of increased adiposity on both A1c and the risk of T2DM with the effect of the same 1-unit increment of increased adiposity later in life to make inferences about optimal timing of weight loss to prevent T2DM. Adiposity was determined as weight, BMI, and waist circumference at time of study enrollment. Finally, the investigators evaluated whether the effect of increased adiposity on A1c is modified by other exposures to infer how to identify those most likely to benefit from weight loss for T2DM prevention. Primary outcome clinically was age at first diagnosis of T2DM and primary biomarker outcome was differences in A1c levels using a time-to-event analysis with proportional hazards models using age as the time-scale (participants censored at T2DM, death, or last follow-up).
Study participants were divided into five quintiles based on PGS, which demonstrated similar mean age (65 years old) and proportion of female participants (54%). Those in the highest BMI but lowest PGS quintiles demonstrated a 5-fold increased risk of T2DM versus participants in the lowest BMI but highest PGS quintiles (HR: 5.34, 95%CI: 5.12-5.53): this suggests that most cases of T2DM are acquired rather than inherited, and therefore preventable. When analyzing random allocation of lifelong exposure to 1-increment of increased adiposity on both A1c and BMI, both had the same effect later in life (A1c: p=0.34; T2DM risk: p=0.67), suggesting weight gain does not cause accumulating irreversible morbidity, but instead, has a reversible morbid effect on both A1c and T2DM risk. There were substantial variations in the effect of BMI on A1c depending on sex, waist circumference (marker of adiposity distribution), ethnicity, and PGS (p<0.001 for all): this variation suggests everyone may have a “personal weight threshold” for developing dysglycemia and ultimately T2DM.
The investigators note the implication of their results and subsequent conclusions on real-world management of T2DM prevention. Given the natural progression of normoglycemia to dysglycemia to T2DM, they propose the possibility of regularly monitoring weight, BMI, waist circumference, as well as A1c in order to determine those at highest risk of developing T2DM. Furthermore, given the effect of adiposity regardless of BMI, they also propose targeting weight loss interventions to those with higher increases in A1c due to weight gain or increased adiposity, regardless of BMI. They suggest such an approach could help target effective prevention to those more likely to develop T2DM, despite lower BMIs, while avoiding overtreatment of those less likely to develop T2DM despite higher BMIs. Ultimately, their results present clinical strategies for targeting prevention of dysglycemia to those most likely to benefit.