Search

OBJECTIVE To evaluate anthropometric, haemodynamic and biochemical risk markers for coronary heart disease (CHD) in the Polynesian population of Western Samoa in a case-control study of 43 cases of CHD compared with 90 age- (mean 53 years) and sex-matched controls. METHODS Cases were identified on the basis of a 12-lead electrocardiogram and clinical history. RESULTS More than 60% of the participants had a body mass index > or =30 kg/m2 and nearly 80% had central obesity. Both diabetes mellitus (17%) and impaired glucose tolerance (9%) were also common in this population. Nineteen per cent of the population were hypertensive and both antihypertensive therapy (21 versus 1%, P

BACKGROUND Asian Indians show an increased risk of non-insulin-dependent diabetes mellitus (NIDDM) and coronary heart disease, together with adverse fat distribution and hyperinsulinaemia relative to other ethnic groups. Using population-based data, we investigated the question of whether the adverse fat distribution observed in Indians can be explained by differences in behavioural risk factor levels. We have examined the question of whether ethnic differences in fat distribution are responsible for the unfavourable risk factor profile of Indians. SUBJECTS AND METHODS Fat distribution (waist: hip ratio) was compared in population-based samples of Asian Indian (n = 4394), Creole (n = 1746), and Chinese (n = 425) Mauritians, after controlling for body mass index and other factors. The contribution of this ratio to ethnic differences in cardiovascular disease risk factors and the role of fasting insulin concentrations were also determined. RESULTS Indian men had the highest mean waist: hip ratio, despite having the lowest body mass index. In Indian women the mean waist: hip ratio and body mass index were intermediate between those of Chinese and Creole women. Indians of both sexes had low levels of high-density lipoprotein (HDL) cholesterol and HDL: total cholesterol compared with Creoles or Chinese, whereas triglycerides levels were highest in Indian men but intermediate in Indian women. Mean fasting and 2 h insulin concentrations were not consistently highest in the Indian subgroups. Blood pressure and serum urate levels were lowest in Indians of both sexes and Indian women also had lower total cholesterol concentrations than either Creoles or Chinese. The elevated waist: hip ratio in Indians was not explained by differences in physical activity, cigarette smoking, or alcohol consumption and the differences in this ratio (and insulin levels) did not explain the observed ethnic differences in metabolic parameters. CONCLUSIONS The susceptibility of Indians to abdominal obesity contributed to the less desirable levels of some, but not all, cardiovascular risk factors. Ethnic differences in cardiovascular risk factors in Mauritians were not explained by differences in abdominal obesity, serum insulin or behavioural risk factors. These data suggest that it is over-simplistic to ascribe the adverse cardiovascular risk factor profile commonly observed in Asian Indians to a tendency to abdominal obesity.

The association of obesity with mortality was investigated in population-based samples of Micronesian Nauruans (n = 1,400), Melanesian Fijians (n = 1,279), and Indian Fijians (n = 1,182), over 10 years from 1982 in Nauru, and 11 years from 1980 in Fiji. At the end of follow-up, vital status was known for all Nauruans and all but 3.5% of Fijians. Mortality rates were higher in Nauru than Fiji, and in Melanesians than Indians. The mean body mass index of decedents was similar to or less than (Nauruan men, p < 0.001) that of survivors in each sex-ethnic group. Crude mortality rates showed an inverse relation with body mass index in Nauruan men, with inconsistent relations in other sex-ethnic groups. After stratification by diabetes status, there was no relation between mortality and obesity in nondiabetic subjects, but an inverse relation was observed among diabetic subjects in each population. These findings persisted even after the exclusion of subjects who died within the first 2 years of follow-up. After controlling for age, smoking, and diabetes status in Cox proportional hazard models, body mass index (as a continuous variable) was not related to mortality in any sex/ethnic group and tended to be negatively associated with mortality risk. Interactions of body mass index with age, smoking, and diabetes status were not significant. Mortality risk was significantly increased in older subjects and in diabetic subjects, and cigarette smoking also increased risk in some groups. Stratification of analyses according to cigarette smoking did not alter the nature of the results. The association of mortality and body mass index categorized by quartiles was also investigated. After adjusting for age alone, or age, smoking, and diabetes status, the lower quartiles of body mass index were consistently associated with the highest relative risk for mortality. Quadratic terms for body mass index did not improve Cox models in subjects with normal glucose tolerance. Relations with cardiovascular disease mortality were also assessed and results were inconsistent, although positive trends were observed in Nauruan women (p = 0.02) and Melanesian men (p = 0.06). Overall, there was little evidence to suggest that obesity was a risk factor for total or cardiovascular mortality in these populations. However, obesity is clearly associated with a high risk of diabetes and other morbid conditions and at least on this basis it would seem desirable to prevent obesity in these and other Pacific populations.

In recent years, developing populations such as the Pacific island nation of Fiji, have seen decreases in infectious diseases and increasing frequency of cardiovascular diseases (CVD), diabetes and cancer. However, cohort studies of mortality in these populations are scarce. Here we report 11-year all-cause and cause-specific mortality rates and risk factors for total, CVD and coronary heart disease (CHD) for indigenous Melanesian and Asian Indian people of Fiji.Following a baseline risk factor survey in 1980, mortality surveillance continue until 1991 in a representative cohort of 1325 Melanesians and 1221 Indians from urban and rural areas of Fiji. Date and cause of death were recorded and total, CVD and CHD mortality rates calculated. Baseline predictors of mortality were assessed using Cox regression.Total mortality rates in Melanesians were 15.9 and 9.2/1000 person-years, and in Indians were 13.5 and 6.8/1000 person-years, in men and women respectively. Death due to CHD was more common in men than women, and in Indian than Melanesian men, although total CVD deaths were more common in Melanesian men. Deaths due to CHD were more common in the urban than the rural area. After adjusting for other risk factors Indian ethnicity was associated with a significantly reduced risk of total and CVD mortality in men, and total mortality in women. Age and systolic blood pressure were consistently and independently associated with mortality from all causes, as well as CVD and CHD (except in Indian women). In men associations were also identified for total cholesterol with CVD and CHD mortality in Melanesians, and 2-hour plasma glucose with total and CVD mortality in Indians. In women, 2-hour glucose was important for total, CVD and CHD mortality in both ethnic groups as was smoking in Indians. Obesity had inconsistent associations with mortality.Cardiovascular disease is now responsible for a large proportion of total mortality in both Indian and Melanesian Fijians. The major risk factors identified in Fijians are similar to those observed in developed populations.

OBJECTIVE To determine the prevalence of diabetic retinopathy and nephropathy retinopathy and nephropathy and to define associated risk factors in Polynesian Western Samoans with non-insulin-dependent diabetes mellitus (NIDDM) or impaired glucose tolerance (IGT). RESEARCH DESIGN AND METHODS A 1991 population-based study in Samoan adults (ages 25–74 years) included a 75-g oral glucose tolerance test, anthropometric measurements, and blood pressure recordings. Subjects with NIDDM or IGT had 45-degree stereo photographs taken (n = 263) (three standard fields of the right eye), and retinopathy was graded in comparison with Airlie House photographs. First-morning urine samples (n = 304) were also collected from these subjects and from a subsample with normal glucose tolerance. Urinary albumin concentration (UAC) was measured by radio-immunoassay: microalbuminuria was defined as UAC of 30–299 μg/ml; and macroalbuminuria was defined as UAC ≥300 μg/ml. RESULTS The prevalence of diabetic retinopathy was 43.2% among subjects with known diabetes and 15.4% in those newly diagnosed. Proliferative diabetic retinopathy was found in 4.5% of known diabetic subjects. The prevalence of elevated UAC was 15.0% in subjects with IGT, 26.0% in newly diagnosed diabetes subjects, and 23.4% in known diabetes subjects. For all diabetic subjects (n = 162), the factors independently associated with diabetic retinopathy (logistic regression) were duration of diabetes, fasting plasma glucose, and body mass index (inversely). Duration of diabetes, serum triglyceride concentrations, and systolic blood pressure were independently associated with elevated UAC in all diabetic subjects (n = 138), and fasting plasma glucose had borderline significance. CONCLUSIONS Diabetic retinopathy and albuminuria are common in Polynesian Western Samoans. Duration of diabetes and level of glycemia were the most important associated factors. These data underline the need for cost-effective programs for the detection and early treatment of diabetes in Western Samoa and other developing populations with high susceptibility to NIDDM.

Cross-sectional associations between insulin and ECG abnormalities suggestive of ‘possible’ and ‘probable’ coronary heart disease (CHD) in the populations of Nauru (n = 568) and Mauritius (n = 3280) have been examined in both non-diabetic and diabetic subjects. Additionally, the longitudinal relationship between baseline insulin and incident ECG abnormalities has been explored in non-diabetic Nauruans (n = 177) over 5 years. Age-adjusted mean 2-hour serum insulin was generally higher in subjects with ECG abnormalities than those with a normal ECG, but the difference was significant only for non-diabetic Mauritian men (p < 0.01). There was no clear association between prevalence of ECG abnormalities and quintiles of fasting or 2-hr insulin in Mauritians, and in non-diabetic Nauruans there was a non-significant positive association between prevalence of ECG abnormalities and tertiles of 2-hr insulin. Logistic regression analyses showed a slight positive association between 2-hr insulin and ECG abnormalities in non-diabetic Mauritians (p = 0.06 in males, p = 0.09 in females), and non-diabetic male Nauruans (p = 0.054) independent of possible confounders. Fasting insulin was not associated in any group. In longitudinal analyses in Nauruans there were no significant differences in mean baseline fasting or 2-hr serum insulin between subjects who maintained a normal ECG and those who developed abnormalities consistent with ‘possible CHD’ (there were no changes suggestive of ‘probable CHD’). The incidence of ECG changes suggesting ‘possible CHD’ was slightly higher in the upper tertile of baseline fasting insulin in both sexes and 2-hr insulin in females, but when other factors were accounted for, multiple logistic regression analyses did not support this finding. Data from the populations of Mauritius and Nauru do not support a major role for serum insulin in ECG abnormalities suggestive of CHD.

OBJECTIVES There is conflicting evidence of a link between circulating insulin concentrations and blood pressure. The aim of this study, therefore, was to investigate this relationship in a representative population sample of different ethnic groups. DESIGN A cross-sectional study was performed, using a random cluster sample of 5080 adults aged 25-74 years from the Indian Ocean island of Mauritius. METHODS Glucose tolerance was assessed with a 75-g oral glucose tolerance test. Hypertension and glucose tolerance were classified using World Health Organization criteria. Subjects known to have diabetes before the survey were excluded from these analyses, leaving 3104 Asian Indian, 1256 African-origin Creole and 384 Chinese subjects eligible for this study. RESULTS There were statistically significant linear correlations between fasting and 2-h insulin and mean blood pressure in all sex-ethnic subgroups, except for Chinese men. However, after controlling for age and the body mass index, partial correlations were much reduced and remained significant only for fasting insulin in Indian men (r = 0.07, P < 0.01) and women (r = 0.09, P < 0.001) and 2-h insulin in Chinese women (r = 0.15, P < 0.05). After controlling for age and body mass index, mean fasting and 2-h insulin concentrations were not significantly different between hypertensive and normotensive subjects in any of 36 possible subgroups defined by ethnic group, gender or glucose tolerance (normal, impaired glucose tolerance or newly diagnosed diabetes). In both men and women with normal glucose tolerance (ethnic groups combined) there was no evidence of increasing systolic or diastolic blood pressure across quartiles of either fasting or 2-h serum insulin within any quartile of the body mass index. By contrast, at any given level of serum insulin there was a clear association between an increasing body mass index and increasing blood pressure, suggesting strongly that insulin is not the final pathway by which obesity influences blood pressure. CONCLUSIONS This study of three ethnic groups does not support the hypothesis that the circulating insulin concentration is an independent determinant of blood pressure in populations.

In transgenic mouse models of Huntington disease (HD) environmental enrichment significantly delays disease onset. A questionnaire-based survey of 154 adults with diagnosed HD (mean 4.2 years postdiagnosis) and a known IT15 CAG repeat length, explored whether premorbid lifestyle may relate to age-at-onset (AO). Participants were drawn from HD outpatient clinics in Australia and New Zealand. Premorbid physical, intellectual, and passive activity levels were used to generate scores in the categories of leisure, nonleisure (education, occupation and domestic duties) and total lifestyle. AO was associated with increased CAG repeat length as expected (r = -0.72, P0.001), but also with a lifestyle that included higher levels of passive activity (r = -0.38, P0.001). Multiple linear regression modeling showed lifestyle passivity to be a variable independent of CAG repeat length in predicting AO (R(2) = 0.54, b = -0.22, P = 0.005). Comparison of the mean AO across tertiles of lifestyle passivity scores showed onset 4.6 years (95% CI = 1.3-7.9) later in the least compared with the most passive tertile. CAG repeat length was also shown to predict lifestyle passivity (R(2) = 0.12, b = 1.08, P0.0005). Neither intellectual nor physical activity showed significant relationships to AO or CAG repeat length in this cohort. Our study leads to two conclusions: that a passive lifestyle may be a preclinical expression of HD, and that it actually contributes to the earlier onset of symptoms. Overcoming the tendency to be passive may substantially delay onset of HD. (c) 2010 Movement Disorder Society.

Objective To determine whether a reported positive association between parity and the development of non-insulin-dependent diabetes mellitus (NIDDM) and impaired glucose tolerance (IGT) is reproducible in other populations. Research Design and Methods We investigated the relationship in data from population-based surveys in four Pacific and Indian Ocean island nations. Women ≥ 40 yr of age at the time of the survey, excluding those in whom diabetes developed before 40 yr of age, were included in this study of Micronesians from Nauru (n = 204) and Kiribati (n = 562), Fiji Melanesians (n = 390), Fiji Indians (n = 247), and mixed-ethnic Mauritians (n = 1333). Subjects in each survey underwent a 75-g oral glucose tolerance test, and glucose tolerance status was ascertained with 1985 World Health Organization criteria. Obstetric information and family history of diabetes were determined by interview. Results Age and body mass index (BMI)-adjusted mean parity increased slightly with worsening glucose tolerance in only two groups, decreased in one group, and was inconsistent in the other two (none were statistically significant). We also found an inconsistent relationship between the number of full-term pregnancies and the prevalence of IGT and NIDDM, although in each population, there was a higher prevalence of NIDDM in the highest parity group (≥ 10 pregnancies) compared with the lowest parity group (1-3 pregnancies). In logistic regression analyses accounting for age, BMI, and family history of diabetes, odds ratio estimates for NIDDM and IGT associated with each pregnancy were not significantly greater than unity. Conclusions The results indicate that there is little if any independent association between parity and the development of abnormal glucose tolerance in these populations.

Associations with insulin resistance and islet beta-cell function have been studied cross-sectionally in a population-based sample of 4278 Mauritians comprising Asian Indian, Creole and Chinese subjects. Insulin resistance and beta-cell function were estimated by a computer solved model based on fasting plasma glucose and insulin concentrations. Insulin resistance increased with declining glucose tolerance, whereas beta-cell function was highest in subjects with impaired glucose tolerance (IGT) and lowest in those with non-insulin-dependent diabetes mellitus (NIDDM). Indian subjects had the highest beta-cell function, while ethnic differences in insulin sensitivity were less marked. This may indicate that deranged beta-cell function rather than insulin resistance is the primary determinant of hyperinsulinaemia and glucose intolerance in Asian Indians. beta-Cell function declined with age more sharply than did insulin sensitivity, suggesting that the age-related decline in glucose tolerance is primarily related to loss of beta-cell function. Body mass index, waist/hip ratio, physical inactivity and female sex were independently associated with insulin resistance and beta-cell function. Subjects with a family history of diabetes had increased insulin resistance, irrespective of glucose tolerance. This or similar models may have application in longitudinal population-based studies which seek to determine the relative contributions of insulin resistance and beta-cell function to the aetio-pathogenesis of NIDDM.

The extent to which random plasma glucose levels underestimate the true prevalence of diabetes has been determined in Micronesian Nauruans. In 337 individuals who were screened on the basis of their random plasma glucose levels, the age-standardised prevalence based on a cut-off of 11.1 mmol/l underestimated the population prevalence based on a complete oral glucose tolerance test by 42% in males and 63% in females. At a cut-off level of 7.8 mmol/l the true age-standardised prevalence was underestimated by 16 and 38%, in males and females, respectively. The use of random plasma glucose concentrations to determine the prevalence of diabetes, as currently defined, seems inappropriate. Performing oral glucose tolerance tests on smaller representative population samples should provide more accurate data at less expense than through large-scale screening utilizing random glucose levels.

Many populations of the Pacific and Indian Ocean region have undergone rapid development and modernization of lifestyle in recent decades. Moreover, there has been a concomitant change in the disease profile, such that non-communicable diseases have surpassed infectious diseases as major health problems and causes of mortality. Within this scenario there has been an increase in the frequency and extent of obesity in these populations. Although obesity was not previously unknown, it is now much more prevalent, particularly amongst urban dwellers, and associated health problems are becoming apparent. During the past 15 years we have performed medical surveys on representative community samples in several Pacific Ocean and one Indian Ocean island population. The surveys were designed primarily to assess the prevalence and associated risk factors for non-insulin-dependent diabetes mellitus (NIDDM) and cardiovascular diseases. From these data we have been able to estimate the prevalence of obesity in each population.

The evidence linking insulin to blood pressure is controversial, and results for groups similarly categorized by body mass, glucose tolerance and hypertensive status are often contradictory. We have investigated the relationship in three population-based samples of Micronesian (Nauru), Polynesian (Western Samoa) and Melanesian (New Caledonia) Pacific islanders, who are known to be susceptible to obesity, non-insulin-dependent diabetes mellitus (NIDDM), hyperinsulinaemia and hypertension. After controlling for age and body mass index (BMI), mean fasting and 2-hr (post 75 g glucose) insulin levels were not significantly different between hypertensive and non-hypertensive subjects, in any population or glucose tolerance sub-group, excepting 2-hr insulin in New Caledonians with normal glucose tolerance. Similarly, there were no strong trends for a higher prevalence of hypertension amongst those with insulin levels in the upper quartile of the distribution of each population, although it was apparent in some BMI/glucose tolerance sub-groups of two populations. Multiple linear regression analyses also showed an inconsistent and where present, weak, independent association between insulin and blood pressure in models predicting both systolic and diastolic blood pressure. We therefore conclude that the hypothesis implicating insulin as a major determinant of blood pressure and as the pathophysiological link between obesity, NIDDM and hypertension is not strongly supported either by the literature or the present data.

This study examined the relation between occupation and cardiovascular disease (CVD) risk factors in 2,795 individuals between ages 35 and 54 years from the rapidly developing island nation of Mauritius. Participants attended a 1992 population-based survey of noncommunicable disease (89.1 % response rate). Occupational status, physical activity in the previous year, cigarette smoking, and alcohol consumption were assessed by questionnaire. Anthropometric and metabolic measures included body mass index (kg/m2),waist-to-hip ratio, fasting serum high density lipoprotein cholesterol and low density lipoprotein cholesterol (LDL cholesterol), triglycerides, 2-hour postload plasma glucose and serum insulin concentrations, and blood pressure. In comparison with professional/skilled workers, age-adjusted means of insulin and glucose, LDL cholesterol, triglycerides, and systolic and diastolic blood pressures were significantly (p < 0.05) lower, and the age-adjusted mean for high density lipoprotein cholesterol was significantly higher for unskilled men. In women, risk factors other than LDL cholesterol varied significantly (p < 0.05) across occupational categories, with homemakers tending to have the least favorable profile. Unskilled workers reported significantly more physical activity (p < 0.01), alcohol consumption, and cigarette smoking (men only) (p < 0.05) than did the other groups. Adjustment for multiple covariates revealed an independent association between occupational status and most CVD risk factors, with physical activity attenuating this association. These results elucidate mediating behaviors of CVD risk across occupational categories that could be applied to intervention strategies in Mauritius.

Background. Obesity and non-insulin-dependent diabetes mellitus (NIDDM) have increased in prevalence in Polynesian Western Samoans over the 13-year period 1978-1991, as the population undergoes an 'epidemiological transition'. Methods. We therefore investigated changes in the frequency of dyslipidaemia over the same period in adults aged 25-74 years, and examined factors associated with dyslipidaemia in cross-sectional and longitudinal data. Subjects were drawn from three geographically defined locations representing different degrees of modernization. Results. The age-standardized prevalence of dyslipidaemia increased in each location between 1978 (n = 1197) and 1991 (n = 1748), with the prevalence of hypercholesterolaemia (≥5.5 mmol/l) increasing from 18% to 36% (P < 0.001), and that of hypertriglyceridaemia (≥2.0 mmol/l) increasing from 9% to 15% (P < 0.001) in the capital city, Apia. In 1991 the highest serum concentrations of total, high density lipoprotein (HDL) and calculated low density lipoprotein (LDL) cholesterol were found in Poutasi (intermediate level of modernization), and the highest triglyceride levels in urbanized Apia. Higher levels of body mass index (BMI), waist-hip ratio (WHR), glucose intolerance, fasting insulin concentration, physical inactivity, educational level, and occupational status were all associated with adverse lipid levels in univariate data. Obesity (BMI in women, WHR in men) and survey location were the most important correlates of abnormal lipid levels in logistic regression models. Fasting insulin was also independently associated with high triglyceride levels in men, while in women the increasing levels of fasting insulin were associated with adverse levels of total, LDL and HDL cholesterol, and triglycerides. In longitudinal data (n = 311), lower baseline levels of cholesterol and triglycerides were associated with greater increases in either parameter at follow-up. Elevated fasting insulin and female gender also predicted increasing cholesterol concentrations, and urban residence predicted an increase in triglyceride levels. Conclusions. Current levels of dyslipidaemia in Western Samoa are similar to those observed in developed Western populations, and are increasing rapidly. These findings, considered along with the high prevalence of other cardiovascular disease risk factors in Samoans, including smoking, obesity and NIDDM, suggest that cardiovascular disease will be a major health concern in the future.

Longitudinal changes in serum insulin concentrations in relation to the natural history of glucose intolerance and factors associated with the incidence of NIDDM were studied in 838 nondiabetic Micronesian Nauruans over the 5.1-year period from 1982 to 1987. In 13 individuals who had data at three time-points and who developed NIDDM only at the final test, 2-h insulin levels followed an inverted V-shaped pattern as glucose tolerance declined to NIDDM. Subjects who were normal (n = 651) or had impaired glucose tolerance (IGT) (n = 187) at the 1982 baseline survey were divided into six natural history categories depending on glucose tolerance in 1987. Changes in glucose tolerance were accompanied by changes in mean 2-h insulin concentration that paralleled the inverted V pattern seen in the 13 individuals. Longitudinal changes in fasting insulin were less consistent, but mean levels increased as subjects developed NIDDM. The 5.1-year incidence of NIDDM was strongly related to baseline fasting and 2-h glucose concentrations, but associations with insulin levels were weak and inconsistent. Neither fasting nor 2-h insulin concentrations contributed to logistic regression models predicting deterioration in glucose tolerance, whereas fasting and 2-h glucose levels were included in all models and BMI also predicted deterioration from normal. These data showing sequential changes in insulin concentrations support the β-cell exhaustion theory of NIDDM pathogenesis. However, in contrast to glucose concentrations and obesity, insulin levels are poor predictors of NIDDM risk in Nauruans. This reflects the complexity of interactions with other metabolic markers and the inability of a single examination to characterize the point along the inverted V curve of insulin secretion that an individual has reached.

Surveys conducted in 10 Pacific island populations and in the multiethnic populations of Mauritius and Rodrigues in the Indian Ocean have provided data on the prevalence of obesity, potential etiological factors and medical hazards associated with obesity. The results indicate that the prevalence of obesity (by body mass index (BMI) in some of these populations is among the highest in the world. Obesity related to degree of modernization is more common in urban than in rural locations and tends to be found more often in women. In two populations where longitudinal data were available, there were dramatic increases in prevalence over relatively short time periods. Obesity contributed to the risk of non-insulin-dependent diabetes mellitus (NIDDM) and was associated with other risk factors for cardiovascular disease (CVD) in all populations, but no relationship could be found with total mortality in three ethnic groups for whom data were available. A genetic susceptibility to obesity combined with social pressures that favor high energy intakes and reduced physical activity are believed to be important in these populations. The challenge for the future lies in developing culturally appropriate programs for preventing obesity and thus reducing associated morbidity, while continuing to research its behavioral and genetic determinants.

Trends in the prevalence and incidence of non-insulin-dependent diabetes mellitus and impaired glucose tolerance were studied in Micronesian Nauruans aged 20 years and above, by linking glucose tolerance data collected during population surveys performed in 1975/1976, 1982, and 1987. Based on World Health Organization criteria, the age-standardized prevalence of non-insulin-dependent diabetes mellitus has remained relatively constant (24.0% in 1987), but the prevalence of impaired glucose tolerance has decreased significantly from 21.1% (95% confidence interval (CI) 17.0-25.3) in 1975/1976 to 8.7% (95% CI 7.1-10.3) in 1987. Between the periods 1975/1976-1982 and 1982-1987, the incidence of progression from normal glucose tolerance to either impaired glucose tolerance (incidence rate ratio (IRR) = 0.55, p less than 0.01) or non-insulin-dependent diabetes (IRR = 0.46, p less than 0.05) has decreased dramatically, while progression from impaired glucose tolerance to non-insulin-dependent diabetes has increased (IRR = 1.23). The overall age-standardized incidence of non-insulin-dependent diabetes has declined from 26.2 cases/1,000 person-years in 1975/1976-1982 to 22.5 cases/1,000 person-years in 1982-1987. As there were no changes in the frequency of recognized risk factors, the decline in incidence of glucose intolerance is probably due to the intensity of the epidemic in Nauru, which has already removed a high proportion of the genetically susceptible individuals from the pool with normal glucose tolerance. Coupled with the observations that mortality is higher and fertility lower in diabetic Nauruans across the age range, the observed decline in the incidence and prevalence of non-insulin-dependent diabetes mellitus and impaired glucose tolerance may presage a fall in the population frequency of the diabetic genotype, at least in its more severe form, as might be predicted on the basis of Neel's "thrifty genotype" hypothesis.

The island nation of Mauritius, located in the southwest Indian Ocean, has a high prevalence of non-insulin-dependent diabetes mellitus (NIDDM) among all of its ethnic groups (Hindu and Muslim Indians, African-origin Creoles, and Chinese). These high rates of NIDDM among groups of varying genetic background provide strong support for the importance of environmental components in the etiology of the disease. Research in Mauritius using a simple activity scale has suggested that physical inactivity may be one of these components. The current investigation further examined the association between glucose tolerance and physical activity in middle-aged nondiabetic residents of Mauritius using a more extensive physical activity questionnaire (the Modifiable Activity Questionnaire). Excluding individuals with NIDDM, a statistically significant (P < 0.05) inverse relationship between physical activity and 2-h post-load glucose concentration was found for both males (rho = -0.14) and females (rho = -0.11). Stratifying by ethnic group, similar inverse correlations were observed in Hindu, Creole, and Chinese males, and in Hindu females (P < 0.05), as well as weaker relationships in Muslim males and Creole females (P < 0.10). Total physical activity remained an independent predictor of 2-h post-load glucose concentration after controlling for body mass index, waist-hip ratio, age, and family history of NIDDM. These data are supportive of a potentially important role of physical activity in the prevention of NIDDM in middle-aged inhabitants of Mauritius.

Subclinical elevation of urinary albumin excretion (microalbuminuria) is useful for the detection of individuals at high risk of developing nephropathy in both insulin-dependent and non-insulin-dependent diabetes mellitus. We have evaluated the performance of Micro-Bumintest, a semi-quantitative test for the rapid detection of microalbuminuria. A total of 1186 samples from a population-based survey of the high diabetes prevalence community of Nauru were studied. Compared with a radioimmunoassay method, the Micro-Bumintest had a sensitivity of 97.6%, specificity of 93.2% and overall efficiency of 94.8%, when 40 micrograms/ml was used as the discriminating albumin concentration defining microalbuminuria. Significant differences in radioimmunoassay results between categories of the Micro-Bumintest were found, indicating its usefulness as a semi-quantitative test. Compared with subjects with normal glucose tolerance, there was a statistically significant progressive increase in the prevalence of microalbuminuria as detected by Micro-Bumintest in subjects with impaired glucose tolerance and diabetes mellitus. Micro-Bumintest performs well as a screening test for detection of microalbuminuria not only in the clinical situation, but also in population-based epidemiological studies.