This large population based study used both registry information and genetic markers to identify and stratify diabetes into 5 distinct clusters. They also identified which complication was more prevalent in which cluster.
Cluster 1 was "Severe Autoimmune Diabetes" (SAID), characterised by early onset, relatively low BMI, poor metabolic control, indulin deficiency, positive GAD antibodies, and a markedly higher HbA1c on diagnosis: 30% of this group were in ketoacidosis on diagnosis. This group accounted for 6.4% of patients.
Cluster 2 was named "Severe Insulin Deficient Diabetes" (SIDD). This group accounted for 17.5% of patients in their study, and showed early signs of diabetic retinopathy than the other clusters. This group accounted for 17.5% of patients in their study.
Cluster 3 had a higher prevalence of NAFLD than the other clusters 15.3% of the patients in their study fell into this group.
Cluster 4 (21.6%) which they named "Mild Obese Diabetics (MOD)".
Cluster 5 (39.1) named "Mild Age-Related Diabetes" (MARD) was associated with older age and only modest metabolic derangement.
Complication screenings and other interventions can be more precisely targeted to patients that are at higher risk, which is the goal of precision-medicine.
I think that this study adds to the body of knowledge that we see anecdotally on this forum all the time. Diabetes is a heterogeneous disease - we are all different. As many have said above, the one constant for all of us classified as Type 2 is that our bodies produce too much insulin, and by we can manage this by balancing our carbohydrate intake to what our individual bodies can tolerate.
It also helps explain why the cured/remission/resolved discussion goes around in circles all the time!