Non-insulin-dependent diabetes in Nauruans: A comparison of newly and previously diagnosed cases

Diabetes Research and Clinical Practice, 1 (1985) 179-184 Elsevier

179

DRC 00028

Epidemiological N o t e

Non-insulin-dependent diabetes in Nauruans: A comparison of newly and previously diagnosed cases H. King, P. Zimmet, K. T h o m a and J. Coventry WHO Collaborating Centre for the Epidemiology of Diabetes Mellitus, Lions International Diabetes Institute, Melbourne, Australia

(Received 25 April 1985, accepted 16 July 1985)

Key words: Non-insulin-dependent diabetes; Natural history; Nauru

Summary A comparison of 148 newly diagnosed ('incident') and 202 previously diagnosed ('prevalent') Nauruan diabetics, examined during a population survey in 1982, has permitted cautious inference regarding the natural history of non-insulin-dependent diabetes (NIDDM) in this Micronesian population. As might be expected the results of the comparison do sugge'~t that Nauruan diabetics undergo further deterioration in glucose tolerance, subsequent to diagnosis of the disease. Plasma glucose concentration is higher in prevalent than incident cases (males: fasting - - 12.2 versus 10.4 mmol/i; 2 h - - 18.9 versus 16.3 mmol/l; and females: fasting-- 13.1 versus 9.3 mmol/l; 2 h - - 20.3 versus 15.8 mmol/l) suggesting that present treatment measures may not be effective in this population. There was also some evidence that the metabolic consequences of N I D D M may be greater in female than in male Nauruans. Apart from plasma glucose concentration, of 8 other biological variables examined by univariate and multiple logistic regression analysis, three showed consistent differences between incident and prevalent cases in females, but none were consistently different in males. A relatively small difference in estimates of obesity was observed between incident and prevalent cases, and this was particularly notable in males. A number of potential sources of bias in this study are highlighted, and definitive, longitudinal studies will be required to corroborate these findings.

Introduction A very high prevalence of non-insulin-dependent diabetes (NIDDM) has been described in Nauruans [1] and in certain other non-Caucasoid populations [2,3]. In these studies, N I D D M was diagnosed according to criteria which relied upon an assess-

ment of glucose tolerance, in terms of a subject's blood glucose concentration (fasting and/or after an oral glucose challenge). The high frequency of the disease, thus defined, led naturally to the question of whether such populations suffered from N I D D M in a form comparable with that seen in Caucasoids. Studies of reti-

0168-8227/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)

180

nopathy in Pima Indian [4] and Nauruan [5] diabetics suggest that the prevalence of microvascular disease in these groups is comparable with that found in Caucasoids with NIDDM. However, because most studies of non-Caucasoid diabetics have been crosssectional, there have been few reports of the natural history of N I D D M in such populations. A comparison of incident and prevalent cases of disease, when both are available for study, may provide clues to the natural history of the disease from cross-sectional data. In this report, newly diagnosed ('incident') and previously known ('prevalent') cases of NIDDM, examined during a population survey of Nauruans conducted in 1982, are compared with respect to chosen biological parameters associated with N I D D M in other populations.

Subjects and Methods Nauru is a small island in the Central Pacific, and Nauruans are of Micronesian ancestry. As a result of wealth gained from the export of phosphate, a valuable, natural resource on the island, Nauruans live a non-traditional lifestyle, and consume large quantities of imported food and drink. Levels of habitual physical activity are low and the prevalence of obesity is high. A total population diabetes survey was conducted on Nauru in January 1982. Details of the survey methods, and findings with respect to the prevalence of N I D D M and microvascular disease have been published elsewhere [5,6]. Response to the survey was 83%, and responders were judged to be representative of the total population of Nauruans. Of the 374 diabetics examined during the survey, values for one or more parameters of interest were missing in 24 cases, and these were excluded from the present study. Of the remaining 350 subjects, 148 (42%) were newly diagnosed (incident) and 202 (58%) were previously known (prevalent) cases. The diagnosis of N I D D M was based upon current WHO criteria [7] modified slightly for field survey conditions - - namely a plasma glucose concentration equal to, or exceeding 11.1 mmol/1 (200

mg/dl) 2 h after a 75 g oral glucose challenge, or a past history of NIDDM. Plasma glucose was measured on-site using the Yellow Springs Instrument 23 AM glucose analyser (Yellow Springs Instruments, Ohio, U.S.A.) which uses a glucose oxidase method of estimation.

Results Mean values for selected, continuous anthropometric, physiological and biochemical parameters are compared in incident and prevalent diabetics in Table 1. All variables (except age) were standardized for age by analysis of covariance. Mean age did not differ significantly between new and known cases in either sex, and age-standardization did not change the level of significance of any difference at P < 0.01. The comparisons demonstrate marked differences between the sexes. In males, significant differences were only observed with respect to fasting and 2 h plasma glucose concentration (P < 0.01) and fasting plasma triglyceride concentration (P < 0.05), prevalent cases having the higher mean values. In females, after controlling for age, highly significant differences were observed with respect to fasting and 2 h plasma glucose concentration, and triceps skinfold thickness (P < 0.001) and plasma urea concentration (P < 0.01). Plasma uric acid and systolic blood pressure were also marginally significant (P < 0.05). Prevalent cases had higher mean values for fasting and 2 h plasma glucose concentration, and plasma urea concentration, and lower values for triceps skinfold thickness, plasma uric acid concentration and systolic blood pressure. Median values of the selected variables were also compared, the results being in accord with the comparison of means. The subjects were also subdivided into 3 groups on the basis of duration of known disease (newly diagnosed, duration < 5 years, duration > 5 years) and the selected variables were examined for trend across the 3 groups. None of the non-significant variables in the former analysis assumed signifi-

181 TABLE 1 MEAN a VALUES OF SELECTED VARIABLES IN INCIDENT AND PREVALENT NAURUAN DIABETICS, 1982 Sex/variable

Incident cases

Prevalent cases

Males

(n = 58)

(n = 99)

Age (years) 2 h plasma glucose (mmol/l) Fasting plasma glucose (mmol/l) Body mass index (kg/m2) Triceps skinfold thickness (mm) Plasma cholesterol (mmol/I) Plasma triglycerides (mmol/l) Plasma uric acid (mmol/1) Plasma urea (mmol/l) Plasma creatinine (mmol/l) Systolic blood pressure (mmHg)

45.5 16.3 10.4 33.7 25.2 4.8 1.6 0.4 5.0 0.1 136.2

Females

(n = 90)

(n =

Age (years) 2 h plasma glucose (mmol/l) Fasting plasma glucose (mmol/l) Body mass index (kg/m z) Triceps skinfold thickness (mm) Plasma cholesterol (mmol/1) Plasma triglycerides (mmol/1) Plasma uric acid (mmol/1) Plasma urea (mmol/1) Plasma creatinine (mmol/l) Systolic blood pressure (mmHg)

43.9 15.8 9.3 37.3 34.8 4.9 1.4 0.4 4.5 0.1 139.7

46.9 20.3 13.1 35.2 30.5 5.2 1.6 0.3 5.6 0.1 132.0

a * ** ***

** **

*

*** *** ***

* ** *

49.1 18.9 12.2 33.1 23.0 4.9 2.1 0.4 5.7 0.I 132.6 103)

All variables (other than age) standardized for age by analysis of covariance. P < 0.05. P < 0.01. P < 0.001.

c a n c e w h e n tested for t r e n d in this way. In a l m o s t all cases, differences were g r e a t e r a n d m o r e c o n sistent b e t w e e n the i n c i d e n t cases a n d t h o s e with k n o w n d u r a t i o n < 5 years, t h a n b e t w e e n the l a t t e r g r o u p a n d t h o s e with d u r a t i o n > 5 years. Coefficients o f l i n e a r c o r r e l a t i o n b e t w e e n variables s h o w i n g significant differences were e x a m ined. W i t h the e x c e p t i o n o f the 2 e s t i m a t e s o f plasm a glucose c o n c e n t r a t i o n , w h i c h were highly correlated, the o n l y c o r r e l a t i o n s > 0.2 were b e t w e e n fasting p l a s m a glucose a n d triglyceride c o n c e n t r a tions, a n d b e t w e e n systolic b l o o d p r e s s u r e a n d triceps s k i n f o l d thickness. T h e d a t a were f u r t h e r e x p l o r e d b y the use o f the m u l t i p l e logistic regression m o d e l [8] p r e d i c t i n g dia-

betic status (incident case, 0; p r e v a l e n t case, 1). A s fasting a n d 2 h p l a s m a glucose c o n c e n t r a t i o n s were k n o w n to be highly i n t e r c o r r e l a t e d , o n l y the latter was tested in the regression models. C o n f o u n d i n g by age was c o n t r o l l e d for by a line a r a n d q u a d r a t i c term. P r e d i c t o r v a r i a b l e s were then selected b y the f o r w a r d selection p r o c e d u r e . T h e m o d e l s c o n f i r m e d the results o f the u n i v a r i a t e analyses. In males, n o v a r i a b l e m a d e a significant i m p r o v e m e n t after the a d d i t i o n o f 2 h p l a s m a glucose c o n c e n t r a t i o n . In females, the m o s t p o w e r f u l p r e d i c t o r v a r i a b l e s were - - in o r d e r o f their e n t r y into the m o d e l - - 2 h p l a s m a glucose c o n c e n t r a t i o n , triceps s k i n f o l d thickness, p l a s m a u r e a c o n c e n t r a tion a n d p l a s m a uric acid c o n c e n t r a t i o n . T h e

182

0 ~d

oo-~%

t~

0

o fL) J

0

I

II

r~

. 0m

< II

0

0 ,.o

I0

,M

I

II

e~

Z L6 0 0

0

~

1

1

,,~ 0

Iu 0

1= 0

< r~

.,o

Z <

r.~

5

I

;~

N ,..1 < > e~

0 r"

z

I

0

~.~

e~ <

~x~ o ~ d l t I

,.0

.o

0

<

[

~4 I o

r~

c~

~.~

o U.I

o~

>

N

~vC~ ~.'~v v

183

parameter estimates, their standardized values, and their order of entry into the model are shown in Table 2 (variables which were not significant in the model for either sex are not included in the table).

Discussion

Nauruans are exceptional for their high prevalence of NIDDM. Although the prevalence of retinopathy in Nauruan diabetics is known to be similar to that expected in other diabetic populations [5] and nephropathy is also known to occur (Zimmet, unpublished data) the natural history of NIDDM in Nauruans remains, to a large extent, a matter for speculation. The subject is of more than academic interest, as the severity with which the disease manifests itself in this population will have important implications for public health planning in Nauru. Clearly the more severe the disease, the more urgent are both primary and secondary prevention measures. The results of this study suggest that Nauruans undergo further metabolic deterioration subsequent to their initial decompensation. Whilst such a result may have been anticipated from extrapolation of information gained from other populations, this aspect of the disease process has not been formally assessed in any Pacific population before. That glucose tolerance may continue to decline after the onset of NIDDM - - either as a result of deteriorating beta-cell function, increasing insulin resistance, or both - - would support the concept of NIDDM in Nauruans as a progressive disease of similar natural history to that observed in Caucasoids, and other more closely studied populations. More difficult to interpret, is the suggestion that the metabolic consequences of NIDDM are greater in Nauruan females than males. This finding is not supported by studies of diabetic retinopathy in Nauruans [5] and Oklahoma Indians [9]. However, although West [10] in an extensive review, cited many studies in which microvascular disease appeared to affect male and female diabetics equally, that author also quoted several in which an excess morbidity was apparent in females. Relative mor-

tality has also been shown to be higher in female than in male diabetics [11] and this has been ascribed to loss of hormonal protection against cardiovascular disease. Clustering of risk factors for cardiovascular disease has also been shown to be more marked in female than in male diabetics [12]. The findings with respect to plasma urea and uric acid in females may well indicate a degree of renal damage in the prevalent group. The relatively minor difference in estimates of obesity between incident and prevalent groups, particularly evident in males, is also of note. This may imply that weight loss occurs very soon after onset of disease, that severity of disease does not affect degree of weight loss, or perhaps that weight loss is not an inevitable consequence of the disease. Of potential sources of bias in this study, perhaps the most serious would be the tendency for more severely affected subjects to enter the prevalent cohort, by virtue of the need to seek medical attention. It is most unlikely such a selection bias was not operating to some extent, and its effect is difficult, if not impossible to determine. However, many studies, including one of Nauruans [5] have shown that duration of disease is the most powerful predictor of diabetic complications, and there is therefore little doubt that the 2 groups compared in this study differ with respect to duration of disease, the factor of interest in this study. As has been reported elsewhere [6], the prevalence of retinopathy was 6% in the incident group, as compared with 37% in the prevalent cases. The effect of treatment may be expected to have introduced a conservative bias upon the results, which showed higher estimates of glucose intolerance in the prevalent group. Bias due to treatment is likely to have been small in the present study, as Nauruans are not generally compliant with treatment. In conclusion, the natural history of NIDDM in Nauruans appears to involve deterioration in glucose tolerance subsequent to the development of disease. The consequences of hyperglycaemia may be greater in females than in males. These findings require corroboration by definitive, longitudinal studies.

184

Acknowledgements We are grateful to the Minister for Health, Republic of Nauru for permission to undertake this study. The analysis was carried out with the support of NIH Grant 1 ROI AM 25446. We would also like to thank the Department of Health, Nauru and the Board of Management of the Royal Southern Memorial Hospital for providing staff and equipment to assist with the survey.

References 1 Zimmet, P., Guinea, A., Guthrie, W., Taft, P. and Thoma, K., The high prevalence of diabetes mellitus on a Central Pacific island, Diabetologia, 13: I 11-115, 1977. 2 Bennett, P.H., Rushforth, N.B., Miller, M. and Le Compte, P.M., Epidemiological studies of diabetes in the Pima Indians, Recent Prog. Horm. Res., 32: 333-376, 1976. 3 Zimmet, P., Taylor, R., Ram, P., King, H., Sloman, G., Raper, L.R. and Hunt, D., The prevalence of diabetes and impaired glucose tolerance in the biracial (Melanesian and Indian) population of Fiji - - a rural-urban comparison, Am. J. Epidemiol.. 118: 673-688, 1983.

4 Dorf, A., Ballintine, E.J., Bennett, P.H. and Miller, M., Retinopathy in Pima Indians, Diabetes, 25: 554--560, 1976. 5 King, H., Balkau, B., Zimmet, P., Taylor, R., Raper, L.R., Borger, J. and Heriot, W., Diabetic retinopathy in Nauruans, Am. J. Epidemiol., 117: 659-667, 1983. 6 Zimmet, P., King, H., Taylor, R., Raper, L.R., Balkau, B., Borger, J., Heriot, W. and Thoma, K., The high prevalence of diabetes mellitus, impaired glucose tolerance and diabetic retinopathy in Nauru - - The 1982 Survey, Diabetes Res., 1: 13-18, 1984. 7 World Health Organization Expert Committee on Diabetes Mellitus, Second Report, Technical Report Series 646, WHO, Geneva, 1980, pp. 8-12. 8 Baker, R.J. and Nelder, J.A., The GLIM system release 3, generalised linear interactive modelling, Numerical Algorithms Group, Oxford, 1978. 9 West, K.M., Erdreich, L.J. and Stober, J.A., A detailed study of risk factors for retinopathy and nephropathy in diabetes, Diabetes, 20: 501-508, 1980. 10 West, K.M., Epidemiology of Diabetes and lts Vascular Lesions, 1st ed., Elsevier, New York, 1978. 11 O'Sullivan, J.B. and Mahan, C.M., Mortality related to diabetes and blood glucose levels in a community study, Am. J. Epidemiol., 116: 678-684, 1982. 12 Wingard, D.L., Barrett-Connor, E., Criqui, M.H. and Suarez, L., Clustering of heart disease risk factors in diabetic compared to nondiabetic adults, Am. J. Epidemiol., 117: 19-26, 1983.