Exposure to human enteroviruses has been implicated as an environmental factor that could trigger and accelerate the autoimmunity leading to type 1A diabetes. In Cuban studies, the presence of enterovirus RNA in sera from newly diagnosed patients with type 1 diabetes was significantly higher than in healthy controls (27 % vs. 3 %) [5, 6]. Also, enterovirus RNA was detected in sera of first-degree relatives of islet cell antibody-positive patients compared to healthy controls (16 % vs. 0 %). Molecular mimicry or the hygiene hypothesis are plausible explanations for these phenomena similar to other populations and regions.

The human T lymphotropic virus type 1 (HTLV-I) virus is endemic in the Caribbean with a prevalence of 2–5 % [7]. While endocrine disorders such as hypercalcaemia have been reported with HTLV-I, there is relatively little data about hyperglycaemia. One study [8] found a higher than expected seroprevalence of HTLV-I in Jamaicans with type 1 diabetes (i.e. 17 %). However, this initial finding is insufficient to prove a causal association and needs further study. Interestingly, HTLV-I virions use the glucose transporter type 1 (GLUT1) to infect CD4(+) lymphocytes [9].

In 1955, Hugh-Jones published one of the first descriptions of diabetes in the Caribbean where he observed that type 2 diabetes was more common than type 1 diabetes in Jamaica [56]. He also described an unusual variant which he called “J-type” diabetes – “J” standing for Jamaica. In more recent times, there have been other acronyms for J- type such as atypical diabetes, phasic insulin-dependent diabetes, ketosis-prone diabetes, type 3 diabetes, ketosis-prone type 2 diabetes and Flatbush diabetes. This variant was associated with insulin resistance, phasic dependency of insulin and a lean phenotype. It most resembles type 1B diabetes in the WHO classification. It is more common in non-white populations, and marginal nutritional status may play a role in some persons. In the Caribbean, there seems to be fewer cases of AKPD conceivably because improved nutrition in the Caribbean over the past several decades is making undernutrition less common. At present, Caribbean clinics are swamped with classical type 2 cases.

Patients with AKPD have periods in which they are insulin-requiring resulting in ketosis, especially during metabolic stresses, e.g. infections [57]. At other times, their need for insulin decreases such that reasonable glycaemic control can be obtained with only lifestyle modification and/or oral antidiabetic agents. Some persons have insulin resistance and others have experienced malnutrition in childhood [58]. It is conceivable that they may have reduced beta-cell mass and glucose-stimulated insulin secretion due to malnutrition in early life similar to the marasmic child [44]. Their beta cells appear to be sensitive to catabolic conditions leading to transient, decreased glucose-stimulated insulin secretion. The precise metabolic triggers (e.g. glucotoxicity, adipocytokines, non-esterified fatty acids causing lipotoxicity) are not known [59].

APKD may be a heterogeneous collection of different phenotypes with different degrees of impaired beta-cell function and autoimmunity (i.e. anti-GAD₆₅ and anti-IA-2 antibodies) [57]. In an African-American series of AKPD which also contained Afro-Caribbean patients, about half had no evidence of autoimmunity (A-) with preserved beta-cell function (β+), while 22 % were A-β-, 17 % were A + β- and 11 % were A + β + [60]. We do not have comparative data in Caribbean persons, but in one study of diabetes in Caribbean youth, who were not necessarily selected on the basis of having ketosis-prone diabetes, 30 % were A-β+, 41 % were A-β- and 39 % were A + β- or A + β + [59, 61]. As such, some persons who have less autoimmunity, i.e. antibody negative but with preserved beta-cell reserve, demonstrate a clinical course more in keeping with type 2 diabetes despite having periods of ketosis [62]. Persons with positive autoantibodies tend to eventually need insulin therapy, while persons with preserved beta-cell function may have periods of insulin independence [57].

A few candidate genes have been examined to explain this variant of diabetes, but no genome-wide association studies have been done to date. A missense mutation Gly574Ser in the transcription factor HNF-1α was thought to be a marker of AKPD in African-American children [63]. However, this candidate mutation was not significant in Afro-Caribbean patients [64]. Other investigators working with other ethnic groups found that variants in HNF-1α and HNF-4α are unlikely to be major contributors to the pathogenesis of type 1B diabetes [65], but this is controversial [66].

In the 1950s and 1960s, diabetes was uncommon in the Caribbean with rates <3 % [69–71]. For example, in 1958, the prevalence was 1.4 % after screening 2325 adults for glycosuria in Trinidad [72] and 0.73 % among 958 Jamaican adults [71]. Notably, women with diabetes outnumber men as far back as 1962 in Trinidad [69], even though surprisingly they were not more overweight than the men.

By the mid-1970s, PAHO sounded the alarm for a looming epidemic of diabetes and NCDs in the region [73]. Since then, the prevalence of diabetes has increased significantly, and the Caribbean, like many developing countries, has an epidemic of obesity and diabetes. The International Diabetes Federation (IDF) estimates that the age-adjusted prevalence in 2013 was 9.6 % for the Caribbean which is the second highest in the world after the Middle East and North Africa region. There is some variation in the rates even within countries, depending on the methods used to diagnose diabetes (fasting glucose vs. oral glucose tolerance testing, 1985 WHO diagnostic criteria vs. the 1997 criteria). Table 8.1 also gives a breakdown by country where logistic regression models were used to produce smoothed age-specific prevalence rates for adults aged 20–79 years, which can allow for country-to-country comparisons. Many countries have not carried out national surveys, and their prevalence rates are extrapolated from other Caribbean countries with a similar demographic profile. Clearly it is important for each country to have their data on the burden of disease.

Naturally, incidence rates of diabetes are also high, although there is some difference based on ethnicity. In Jamaica, which is predominantly of African ancestry, the rates are 15 and 20 per 100 person-years for men and women, respectively [21]. However in Trinidad, where approximately 35 % of the population are of Indian ancestry and another 35 % are of African, the incidence of diabetes is higher in Indo-Trinidadians. The rates of Indo-Trinidadian men were 24 per 1000 person-years compared to Afro-Trinidadian men (13 per 1000 person-years) [17]. This was also similar for Indo-Trinidadian women (23 per 1000 person-years) and Afro-Trinidadian women (14 per 1000 person-years) [17]. On account of the high incidence rates, the IDF has been predicted that the prevalence of diabetes in the Caribbean will increase by 37–60 % over the period by 2035. The epidemic is set to continue as there is a significant reservoir of prediabetes. Impaired fasting glucose occurs in ~3 % [74] and impaired glucose tolerance in 13.7 % [10]. Of note, most surveys did not perform oral glucose tolerance testing, so most of the region’s estimate by the IDF (13.2 %) is actually extrapolated from one late-1990s Jamaican survey [10]. Remarkably, there are little rural-urban differences [74] although there are differences by income status (lower income having higher prevalence). Undiagnosed diabetes is a heavy burden for the region and will increase the morbidity, mortality and loss of human capital. About 25 % of people with diabetes in the Caribbean are unaware of their status [74, 75].

Although cases of type 2 diabetes have been reported in youth, prevalence data is missing in many countries. Anecdotally, these cases have been increasing with the increasing girth of the region’s youth. In the US Virgin Islands, the incidence of type 2 diabetes in youth age ≤19 years rose significantly between 2001 (5.3/100,000) and 2010 (12.5/100,000) giving an age-adjusted annual incidence rates of 9.6 per 100,000 [76]. In Trinidadian schoolchildren, urine screening had a positive predictive value of 65 % for detecting diabetes, and it showed a prevalence of 10·4/100,000 while 7·5/100,000 had IGT [77]. This would, of course, be an underestimate compared to blood screening. In a small Jamaican study, type 2 diabetes occurred in a third of cases of diabetes diagnosed in youth less than age 25 years [78]. Affected youth were more likely to be female, older at diagnosis, obese and have a higher blood pressure when compared to those with type 1 diabetes. However, the strongest predictor of type 2 diabetes was obesity measured by BMI [78].

Like many regions, type 1 diabetes is uncommon in the Caribbean. Precise prevalence and incidence data are sparse in many countries, but data exist for Bahamas and the Eastern Caribbean. The prevalence was very high (31/100,000) in Bahamian youth <age 24 years with an incidence rate of 10.1/100,000 in those age 0–14 years [79]. Incident rates are 4–5 per 100,000 person-years in persons with African ancestry in the Eastern Caribbean [80, 81] rising to a peak in the US Virgin Islands (15.3 per 100,000 person-years) [76, 81] which may reflect their higher degree of genetic admixture. The rates show some seasonal variation and a tendency for a secular increase, which augurs for unknown environmental factors being part of the aetiology such as the accelerator hypothesis [80]. About half of patients with type 1 diabetes are type 1A (with anti-GAD₆₅ and/or IA-2 antibodies positivity), and half are type 1B (i.e. autoantibody negative) [61].

The epidemiologic transition would be expected to increase the risk for gestational diabetes, as well as the number of women who enter pregnancy with type 2 diabetes. About 4–13 % of Jamaican women of child-bearing age have diabetes [10, 74]. There has been an apparent increase in the incidence of GDM in Trinidad. Preliminary data show that from 2005 to 2007, the prevalence of GDM increased from 1.7 to 6.7 % with a mean prevalence of 4.3 % [82]. Unknown genetic factors may play a role. A family history of early-onset autosomal-dominant type 2 diabetes seems to increase the risk of GDM in Jamaican women (12 % vs. 1.5 % in controls, i.e. OR 9.0) which would have implications for screening [83].

Women with prior GDM are at very high risk for incident type 2 diabetes. So, among Trinidadian women with prior gestational diabetes, 62 % develop diabetes and 17 % had IGT within 7 years [84]. These data suggest that 10–18 % of women convert to diabetes per year, and the conversion to IGT occurs in another 3–5 %/year. If confirmed, these rates are extraordinarily high compared to other regions of the world [84].

People of African ancestry including Afro-Caribbean populations have lower rates of myocardial infarction compared to Caucasians [98, 99], but conversely stroke [100] and renal failure [101] rates are higher. It is not clear why such disparities exist, but genetic factors [98], coexisting cardiometabolic risk factors [100], access to care and intensity of therapeutic control of risk factors are potential culprits. There is limited data on genetic factors influencing complications in diabetic Caribbean people. Persons of African ancestry have higher systolic blood pressures, but lower triglyceride and total cholesterol levels [99]. Hypertension occurs in 52 % and 35 % of diabetic Jamaican women and men, respectively [10], and there may be less nocturnal dipping of blood pressure [102]. Tobacco use is relatively low in diabetic persons [10] compared to developed countries. As a result, these co-morbid cardiovascular differences may explain some of the higher rates of stroke and renal failure (which are sensitive to blood pressure), and the less atherogenic lipid profile may reduce the risk of coronary artery disease. Notably, youth with diabetes have a more adverse cardiometabolic pattern and thus may be more at risk of complications [103], as was also seen in the SEARCH study.

The leading cause of death in the Caribbean is ischemic heart disease, and stroke is the second of which diabetes is a major cause [104]. Also, diabetes is the leading causes of disability-adjusted life years. Mortality rates from diabetes have increased dramatically, i.e. 63 % from 1980 to 1990. Diabetes-attributable macrovascular complications are affected by ethnicity. The population-attributable mortality of Indo-Trinidadians is 2.9–6.9 times higher than other ethnic groups [98, 105], and most of this is due to diabetes-induced cardiovascular disease. The age-adjusted death rates due to diabetes in 2000 were 25, 46, 56, 58 and 108 per 100,000 world standard population for Suriname, Bahamas, Barbados, Jamaica and Trinidad, respectively, according to PAHO (Health Statistics for the Americas, 2006 Edition, PAHO). In North America, the rates are less than 16 per 100,000, so these data show significant health disparities. In Barbados, diabetes accounted for an excess mortality of 42 %, and there was a 9 % increase in all-cause mortality for each 1 % increase in A1c [91]. However, there is dearth in data in other Caribbean countries about specific macrovascular complications and their mortality. Cardiovascular disease is present in almost 60 % of hospitalised diabetic Jamaicans and is more frequent among women [106]. Silent MI may occur in a quarter of Guadeloupian patients especially if they have left ventricular hypertrophy [107]. The atrial natriuretic peptide rs5065 (2,238T>C) C allele seems to exert a protective effect (24 % vs. 41 %, OR 0.5) in a relatively small study in Guadeloupe [108].

Diabetes-related mortality is not limited to older age groups, as 38 % of deaths occurred in people under the age of 60 [109, 110]. Mortality in people with type 1 diabetes in the US Virgin Islands is high, as cumulative survival was 98 % at 10 years, but fell to 73 % at 20 years. This high incidence rate of T1D in the US Virgin Islands may be partially responsible for the high mortality rate also seen [111].

Diabetic foot disease has major public health consequences for the region. Approximately one of every eight patients in diabetes clinics had a major foot complication (amputation, ulcers, infection). Factors associated with these complications were neuropathy (OR 9.3), high blood pressure (OR 7.9) and longer duration of diabetes (OR 1.32) [112]. Diabetic foot disease accounted for 30 % of admissions in Barbados and 89 % of diabetes-related admissions [113]. In fact, Trinidad uses 0.4 % of their gross domestic product solely to treat patients hospitalised for diabetic foot infections [114].