The syndrome is poorly understood, with little information available on the prevalence, aetiology, triggers and overall pathogenesis.
Of the properly researched cases incl control group populations, a hi/normal carb diet, corresponding high insulin regime, BMI <28, age < 50 yo and male gender, appear to be the only common observations among those with diabetes having been declared dead with the Sudden Unexpected Death Syndrome (SUDS). Mind you, not all have been found dead in their beds... So not necessarily 'young' of age, and also not just the recently new diagnosed with diabetes either. As you mention in discussion above, the sample size is really small, reason why findings are not statistically significant either as far as I know. Also non-diabetics are found dead at times with no obvious reason found at autopsy.
With regards to the diabetics ending this way (and its still low on the frequency list compared to other causes of death), the following has been considered:
Acute hypoglycaemia may play a role. However, going hypo is a common occurrence among patients with type 1 diabetes, but it rarely results in sudden death. This implies that other mechanisms are likely to be involved. E.g. patients with long-standing type 1 diabetes may have reduced parasympathetic activity and increased sympathetic activity. These effects have been associated with lowered baroreceptor-cardiac reflex sensitivity in individuals with type 1 diabetes, which could result in increased susceptibility to ventricular arrhythmias. (which is actually attributed to presumed main cause of death for SUDS in non-diabetics population). Moreover, the risk of such ventricular arrhythmia is compounded by nocturnal episodes of hypoglycaemia. Research has demonstrated that patients with type 1 diabetes, with normal or abnormal cardiac autonomic function, have a prolonged QTc (corrected QT) interval during a hypo. Genetic predisposition to prolonged QTc interval prolongation is well recognised in the general community. Considering this, it is most probable that multiple factors in one person — T1, a genetic predisposition to QT interval prolongation, cardiac autonomic neuropathy and acute severe hypoglycaemia — collectively cause arrhythmia and lead to dead-in-bed syndrome.