Atypical or ketosis-prone type 2 diabetes was first reported by Winter
et al. in 1987 in black Americans[
19] and is distinguishable from other subtypes of diabetes by clinical, immunological, and biological features. It has since been reported in other ethnicities such as Native-Americans, Chinese, and Hispanics[
20‐
22]. Atypical ketosis-prone diabetes is frequently detected in obese individuals and characterised by an acute onset with either ketosis or ketoacidosis[
6,
23]. In addition, patients with atypical ketosis-prone diabetes lack markers associated with islet cell autoimmunity.
The clinical, metabolic, and immunological characteristics of atypical ketosis-prone diabetes have been well studied by other investigators[
4,
21,
22,
24]. However, controversy exists concerning the classification of atypical ketosis-prone diabetes because the disease possesses features associated with both type 1 and type 2 diabetes. Atypical ketosis-prone diabetes was previously classified as idiopathic Type 1 diabetes by the World Health Organization and the American Diabetes Association[
1]. On the other hand, some researchers think that atypical ketosis-prone diabetes is a subtype of type 2 diabetes that develops in patients with high sensitivity to glucotoxicity or lipotoxicity or with dysregulated glucagon secretion[
2,
3,
6]. Furthermore, the prevalence and features of atherosclerosis in both type 1 and type 2 diabetes are well-documented, but those in patients with ketosis-prone type 2 diabetes remain unknown and have not been investigated. Thus, in the current study we focused our attention on investigating the prevalence and clinical characteristics of carotid atherosclerosis in patients with diabetic ketosis but without islet autoantibodies at onset. To the best of our knowledge, this is the first report to describe the prevalence and clinical characteristics of atherosclerosis in ketosis-onset type 2 diabetes.
General clinical features
Because our subjects were negative for islet-associated autoantibodies, the relationship between ketosis-onset and a transient dysfunction of islet cells through an autoimmune process was considered to be unlikely in our studied cohorts. Therefore, we proposed two distinct diabetic groups based on presence or absence of diabetic ketosis. The prevalence of ketosis-prone type 2 diabetes is not known, but our findings indicated that what was once described as “atypical diabetes” is in fact a common clinical phenomenon affecting nearly 50% of Chinese patients with newly diagnosed diabetes in our study. The reason that nearly half of newly diagnosed diabetic subjects had ketosis-onset is that the condition of ketosis-onset diabetic subjects tended to be more serious, and these patients were thus more likely to be hospitalised.
The patients with ketosis-onset diabetes showed an earlier onset, with a mean age of 49 years considering that the mean age of the patients with non-ketotic type 2 diabetes was 56 years old. Most importantly, unlike cases of typical type 1 and type 2 diabetes, which display a similar sex distribution, ketosis-onset diabetes shows a stronger prevalence in males consistently with previous reports[
5,
25‐
27]. Additionally, consistent with several previous studies[
25,
26], we also found that there were higher blood glucose levels and lower plasma C-peptide levels in ketosis-onset diabetics compared with the non-ketotic type 2 diabetics, which indicated the insulin secretion function of islets was impaired at onset in the patients with ketosis-prone diabetes. The acute impairment of β-cell function in ketosis-prone diabetes has been found to be the major determinant of ketosis onset[
24]. Although the underlying mechanisms of impaired islet β-cell function have not been elucidated, an increased susceptibility to glucose toxicity or lipotoxicity might result in the transient functional abnormalities of islet β-cells[
25,
27,
28].
On the other hand, there were many similarities in the clinical characteristics between the ketosis-onset diabetics and the non-ketotic type 2 diabetics. For example, the proportion of overweight and obesity in the ketosis-onset diabetics was similar to that of the non-ketotic type 2 diabetics. Obesity was present in 46.6% of the ketosis-onset diabetic patients in our study, which was very close to the findings of Winter and colleagues[
19]. Additionally, both BMI and WHR were also similar between the two groups, which indicated that individuals with ketosis-prone diabetes physically resemble those with type 2 diabetes. Likewise, we observed older age, a higher proportion of hypertension, higher SBP, and a greater mean CIMT value in subjects with carotid atherosclerosis than in those without carotid atherosclerosis, and this was shared by between ketosis-onset and non-ketotic type 2 diabetic groups.
Carotid atherosclerotic lesions
Previous studies have described the prevalence of carotid atherosclerosis in both type 1 and type 2 diabetes, but there are very few reports available for ketosis-prone type 2 diabetes. Thus, we further evaluated the prevalence and clinical features of carotid atherosclerosis in ketosis-onset diabetes. Our results demonstrated a significantly increased prevalence of carotid atherosclerosis in newly diagnosed diabetic patients with ketosis-onset compared with the control subjects without diabetes but not in comparison to the non-ketotic diabetic patients. Moreover, the diabetic patients with ketosis-onset had a nearly 2.3-fold increased risk of carotid atherosclerosis in comparison to the control subjects consistent with findings of other studies of typical type 2 diabetic patients. On the contrary, when compared with the non-ketotic type 2 diabetes, the risk of carotid atherosclerosis was not markedly increased or decreased in the ketosis-onset diabetes. In a population-based study, carotid atherosclerosis was present in 58% of the study subjects[
29]. A study by Lundman
et al.[
30] showed that 28% of type 2 diabetic patients had one or more carotid plaques. After a 2-year follow-up, the prevalence of carotid plaques increased to 62%, including patients who initially had plaques. It is well-established that carotid artery plaques are markers of systemic subclinical atherosclerosis and strong predictors of cardiovascular events[
29,
31,
32]. A large population-based study has demonstrated a significant association between the presence of carotid plaques and the risk of vascular events[
32]. Similar to other studies[
33‐
35], the prevalence of carotid atherosclerosis also significantly increased with age in the patients with ketosis-onset diabetes, but there was no significant difference between sexes in our study which may be due to uneven gender distribution in diabetic patients with ketosis-onset[
5,
25,
27]. Given that carotid atherosclerosis is associated with an increased risk of cardiovascular disease and is also a powerful predictor for future cardiovascular events, the presence of carotid atherosclerosis in ketosis-prone diabetic patients may contribute to the improved prediction of the future risk of vascular events.
In contrast to carotid atherosclerosis, the prevalence of carotid stenosis was low in both newly diagnosed ketosis-onset diabetics and non-ketotic type 2 diabetics, and no significant difference was noted among the three groups studied. The estimated prevalence of carotid artery stenosis varies to a great extent in different studies and is dependent on the definition of carotid stenosis, the method for measuring stenosis, and the selected study populations[
13,
36,
37]. In the Tromsø Study, the prevalence of carotid stenosis was 5.3% in men and 3.8% in women in the general population[
36]. Lacroix
et al. demonstrated that the prevalence of carotid stenosis <60% and ≥ 60% was 63.6% and 4.7% in type 2 diabetic patients without any history of cerebrovascular disease, respectively[
37], which was far higher than that of our study subjects. This discrepancy was partly explained by the fact that our study subjects were newly diagnosed with diabetes although atherosclerotic stenosis is more frequently detected in the diabetic patients with a long history of diabetes. Furthermore, the prevalence of carotid stenosis is significantly higher in men than in women and increases with age as shown in other studies[
36,
37]. In our study, although the prevalence of carotid stenosis stratified by sex and age was not statistically significant different in the patients with ketosis-onset diabetes, the prevalence of carotid stenosis was also higher in men (3.9%) than in women (0.0%). Likewise, the prevalence of carotid stenosis increased with age in patients with ketosis-onset diabetes. Given that ketosis-onset diabetes has a strong male predominance, the above results may be due to small samples and uneven gender distribution in the ketosis-onset diabetic group. Carotid stenosis is one of the main causes of cerebrovascular events, and the degree of carotid stenosis is also a valid marker of the risk of cerebrovascular events. Therefore, the presence of carotid artery stenosis can be used as an indicator to select patients for interventional procedures among ketosis-prone diabetics, especially in asymptomatic subjects.
Similarly to carotid atherosclerosis, the mean CIMT value showed a significant increase in the ketosis-onset diabetic subjects in comparison to the control subjects. Additionally, in ketosis-onset diabetes the mean CIMT value was directly related to the age of the patients, and the elderly patients had a higher mean CIMT value compared with younger patients consistent with findings made by previous investigators[
16,
38‐
41]. Interestingly, we could not find gender differences in CIMT values in the ketosis-onset diabetic patients, and this was in accordance with a previous study in children with type 1 diabetes mellitus[
40]. We speculate that it may also be due to small sample size and uneven gender distribution in the ketosis-onset diabetic group. The measurement of the IMT of carotid arteries assessed by ultrasound is often used to detect early atherosclerotic lesions and is considered as a surrogate marker of subclinical atherosclerosis. Numerous studies have demonstrated that both type 1 and type 2 diabetic patients had significantly higher CIMT than control subjects. For example, Järvisalo
et al. demonstrated that children with type 1 diabetes had increased carotid IMT compared to control subjects based on ultrasound examination[
8]. Additionally, Pujia
et al. observed that type 2 diabetes subjects had larger CIMT compared with control subjects[
39]. It is well-established that increased CIMT is associated with cardiovascular risk factors such as metabolic syndrome (MS) and insulin resistance (IR), coronary atherosclerosis, and cardiovascular events[
42‐
47]. Herder
et al. found that MS was associated with CIMT and progression of CIMT[
47]. Subjects with MS have higher levels of CIMT at follow up than those without MS[
47], which indicate MS may be involved in the initiation of the process of atherosclerosis. Furthermore, subjects with MS have significantly higher incidence of subclinical carotid atherosclerosis and cardiovascular events than subjects without MS, irrespective of its several definitions[
48]. Obesity is closely associated with IR and MS. However, even in non-obese adolescent type 1 diabetes, CIMT is increased and also associated with IR[
46]. Thus, IR may be used an important factor reflecting early signs of atherosclerosis in type 1 diabetes. We therefore hypothesise that an increase of CIMT in diabetic patients with ketosis-onset may also result in a higher risk of cardiovascular disease in the future.
Consistent with other studies[
33,
49‐
51], we found that some of the traditional risk factors for atherosclerosis were also present in the patients with ketosis-onset diabetes. As expected, age, hypertension, LDL-C, and CIMT were independently associated with the presence of carotid atherosclerosis in the ketosis-onset type 2 diabetes. Therefore, strict control of hypertension and dyslipidaemia is important in order to prevent atherosclerosis in the carotid arteries in diabetic patients with ketosis-onset.