Diabetes Research and Clinical Practice, 20 (1993) 75-85 0 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved. 0168-8227/93/$06.00
Diabetes mellitus and hypertension based on the family history and 2-h postprandial blood sugar in the Ann-Lo district (Northern Taiwan) Jen-Der Lina, We&Bin Shiehb, Miau-Ju Huanga and Hong-So Huanga ‘Division of Endocrinology and Metabolism, and bDepartment of Medicine, Chang Gung Memorial Hospital, Taipei, Taiwan, Republic of China
(Received 6 July 1992; revision accepted 30 November 1992)
From July 1988 to June 1990, we performed an epidemiological study on the prevalence of hypertension and diabetes mellitus (DM) in the Ann-Lo district, a suburban area of Northern Taiwan. One third of the population in the district was randomly sampled. A total of 9087 persons were screened with 67.4% participating. Following completion of the questionnaire blood pressure, 2 h postprandial blood sugar were determined. DM was defined when the blood sugar was over 200 mg/dl or the subject had a diabetic history. Hypertension was defined if the systolic blood pressure was over 160 mmHg or the diastolic blood pressure over 95 mmHg. Statistical comparisons were performed with a chi-square test, analysis of covariance, stepwise multiple regression and Pearson correlation matrix. In this study, the prevalence rate of DM was 2.6% and of hypertension was 6.4%. For those patients 40 years or older, the prevalence rate for DM was 8.0% and for hypertension was 19.7%. The prevalence of hypertension was 24. 8% in overt diabetes and 5.2% in the normal subjects. Analysis of the data between risk factors of DM revealed that DM correlated with age, body mass index, hypertension, smoking, family history of DM and correlated negatively with education. Hypertension correlated with gender, alcohol intake and smoking, sugar level, age and body mass index. The prevalence rate of diabetes and hypertension were quite high in the district and this study pointed out the important risk factors for hypertension and DM in Taiwan.
Key words: Prevalence; Body mass index
Hypertension and diabetes mellitus (DM) are two major chronic diseases which affect the people Correspondence lo: Dr. Jen-Der Lin, Division of Endocrinology and Metabolism, Chang Gung Memorial Hospital, 199 Tung Hwa North Road Taipei, 105 Taiwan, ROC.
in Taiwan and all over the world. These two chronic disorders are the important risk factors for stroke occurrence in Taiwan [1,2]. In the last 10 years, stroke, hypertension and DM were among the ten highest leading causes of death in Taiwan. Epidemiological studies of these chronic diseases were thus quite important in providing us not only with an insight into the incidence, prevalence, risk
factors but also the etiology and complication of these disease. In Taiwan, epidemiological surveys of hypertension and DM have been conducted in subjects over the age of 40 years [3-51, but there is no information concerning the prevalence rate and risk factors for hypertension and DM which cover all age groups in a single district, Although the clinical impression and previous epidemiological studies indicated that hypertension was associated with DM [6,7], we need more convincing data for the Chinese population. The present study was designed to examine the distribution of blood glucose and blood pressure, the prevalence of hypertension and DM, the risk factors for hypertension and DM, and to inform people with regards to the importance of chronic diseases. The Ann-Lo district is located in Keelung city, in Northern Taiwan and consists of people with a suburban life style.
Materials and Methods The present survey was performed between July 1988 and June 1990 in the Ann-Lo district of Keelung in Northern Taiwan. The population density and life pattern of the Ann-Lo district belong to the suburban area. This area was selected as community based epidemiological study due to the area can represent the life style of most Taiwan’s communities. The study included all age groups in the district. One third of all the families in the district were sampled randomly. After an appointment was made, the sampled families were visited door by door, and all the family members were surveyed. A total of 9087 subjects were screened. The subjects included 52.2% females and 47.5% males. The age distribution of the subjects is shown in Table 1 and approximates closely to the age and sex distribution of the Taiwanese population at large in 1988. Adult inhabitants 19 years or older constituted 67.5% of the screened subjects. These subjects represent the main population for analysis of risk factors in the present study. Before the survey was started, we designed a questionnaire which provided suitable information for computer processing. The questionnaire in-
Number diabetic district
and percentage according to different age groups in and hypertensive epidemiological study in Ann-Lo
2953 1654 1836 981 727 654 282
32.5 18.2 20.2 10.8 8.0 7.2 3.1
eluded the following data: file number, name, gender, age, educational level, body height, body weight, habits of drinking and smoking, personal history of diseases, family history of DM and hypertension, blood pressure and two hours postprandial blood glucose or urinary glucose. We asked specially trained nurses or medical students in blood pressure measurement and linger sugar check up to visit the sampled families. After completion of the questionnaire, we checked blood pressures and two hours postprandial blood glucoses with Glucocheck SC B-2  for all subjects 10 years or older. For those children less than 10 years old, we checked the urine glucose first. If this test was positive or trace positive for glucose, the children’s blood glucose were checked. Blood pressures were measured with a mercury sphygmomanometer with a proper sized cuff. Systolic pressure was determined at the first successive sound and the diastolic blood pressure was determined at the disappearance of the sounds. Both systolic and diastolic pressure was recorded to the nearest 2 mmHg. If the subjects received antidiabetic or hypertensive agents, we recorded the initial blood pressure and blood glucose level at the time of diagnosis. Body weight and height were measured at the same time. The educational level of the subjects was separated into 14 levels depending on the duration of the subject’s education. We extended
family histories of PM and hypertension to the second degree relatives. The sampled subjects were not included in the study if they refused an interview or we were unable to meet them after three visits. The total response rate of the sampled subjects was 61.4%. Hypertension was defined according to the WHO criteria: a systolic blood pressure over or equal 160 mmHg or diastolic blood pressure over or equal 95 mmHg or the subjects with a history of hypertension and currently receiving treatment. The two hours postprandial blood glucose over or equal 200 mg/dl or subjects with a diabetic history and receiving treatment were defined as overt DM. The body mass index (BMI) was defined as: weight (kg)/height (m2).
The data in the present study were expressed as mean and one standard deviation. Statistical comparison was performed utilizing chi-square analysis. Matrix of correlation coefficient, analysis of covariance and stepwise multiple linear regression were used to analyze the risk factors related to hypertension and DM in the questionnaires. We used a P-value < 0.01 as criterion of statistical significance.
Results Figure 1 demonstrates the mean and standard deviation of systolic and diastolic blood pressure in the different age groups of all the subjects. As
Systolic Blood Pressure m-
Diastolic Blood Pressure
1 z 80
Fig. 1. Mean systolic
31-40 and diastolic
with S.D.in different
1m= Body k~ass Index
Fig. 2. Mean BS and BMI with S.D. in different age groups.
indicated, systolic blood pressure increased progressively with age. Diastolic blood pressure also increased with age until the age of 61-70 years, when a plateau was reached. The difference between the systolic blood pressures of the different age groups became prominent after the age of 40. The blood glucose level and BMI of different age groups are shown in Fig. 2. Similar to the diastolic blood pressure, blood glucose become elevated with age and reached a plateau at the age of 6 l-70 years. BMI also increased with age, however it reached a plateau earlier than blood glucose levels and blood pressure. Basically, increment of the BMI after the age of 40 is not so salient when compared with the blood pressure and blood glucose. In Table 2, the blood glucose range distribution of all subjects is illustrated. Among them, 86.2% of the subjects had normal postprandial blood sugar
(BS), 11.2% of the subjects had 2 h postprandial BS between 140 and 200 mg/dl, 2.7% of the subjects had 2 h postprandial BS exceeding or equal to 200 mg/dl and were diagnosed as overt diabetics. After age adjustment with the general TABLE 2 Number and percentage of 2 h postprandial BS (BS) ranges in Ann-Lo district Diagnosis
< 140 140 < BS < 200 2 200
7831 1014 242
86.2 11.2 2.7
79 TABLE 3 Number and percentage of different blood pressure groups in Ann-Lo district %
Blood pressure (mmHg)
S.P.a < 160, D.P.a <: 95 S.P. 2 160, D.P. < 95 S.P. > 160, D.P. 2 95 S.P. 2 160, D.P. re 95
8497 32 199 359
93.4 0.4 2.2 4.0
Total hypertension prevalence rate: 6.6%. Age adjusted hypertension prevalence rate: 6.4%. @S.P.,systolic blood pressure; D.P., diastolic blood pressure.
Taiwan population, the prevalence rate of DM in this study was 2.6%. The prevalence of DM in males was 19.59/1000 and in females was 26.16/1000. The male to female ratio for DM was 0.75. Table 3 demonstrates the blood pressure ranges of all surveyed subjects. In the population, 93.4% of the subjects had normal blood pressure, 0.4% of the subjects had systolic hypertension, 2.2% had diastolic hypertension and 4.0% of the subjects had both systolic and diastolic blood pressure above the normal range. In the present study, 6.6% of the subjects were diagnosed as hypertension. After age adjustment, the prevalence of hypertension was 6.4%. In 1000 male subjects the prevalence of hypertension was 121.08 and in the female subjects it was 101.47.
Fig. 3. Prevalence of DM and hypertension in different age groups in Ann-Lo district.
According to age distribution (Fig. 3), we find that the DM and hypertension prevalence rates increased markedly with age. The DM prevalence rate increased from 0.3% for the age group < 19 years to 12.4% for the age group 61-70 years. In the age groups 31-40 years and 41-50 years, the prevalence rate for DM increased by a factor of 2.6. The prevalence of hypertension in this population increased from 0.8% to 27.8%. Similarly, in those subjects in the age groups between 31-40 and 41-50, the prevalence rate of hypertension increased 3.8 times. In the study, 57.3% of the diabetic subjects and 53.4% of the hypertensive subjects were known to have the illness before the survey. The prevalence rates of hypertension and DM in the age group over 40, were 19.6% and 8.0%, respectively. After age adjustment with the population in Taiwan, the prevalence rates were 19.7% and 8.0%, respectively (Table 4). Table 5 illustrates the relationship between DM and hypertension. The prevalence rate of hypertension in the three different BS groups was 24.4%, 11.9% and 6.7%, respectively. The correlation between the level of BS and blood pressure was significant (P < 0.001). In Taiwan, most adult diabetic subjects were type 2 DM . We analyzed the data from questionnaires to study the relationship between the risk factors of hypertension and DM using the matrix of correlation coefficient (Table 6). From this analysis, we found blood glucose to be correlated with age, systolic blood pressure, diastolic blood pressure, BMI, smoking; and negatively correlated with education and a family history of hyperten-
Prevalence of DM and hypertension over 40 in Ann-Lo district
Diabetes mellitus Hypertension
8.0 (8.0a) 19.6 (19.7a)
Total population of age over 40 are 2448. a After adjustment with the population in Taiwan.
and blood pressure
BS (mg/dl) < 140
I40 i BS 5 200
S.P. 2 160 D.P. < 95
S.P. < 160 D.P. b 95
S.P. r 160 D.P. 2 95
S.P. < 16Oa D.P. < 95a
x2 = 2268.8, P < 0.001. a S.P., systolic blood pressure;
sion (P < 0.001). Blood glucose correlated with a diabetic family history too (P < 0.01). Systolic blood pressure correlated with diastolic blood pressure, age, blood glucose level, gender, smoking, drinking, BMI; and negatively correlated with diabetic family history (P < 0.001). Diastolic blood pressure correlated with systolic blood pressure, age, blood glucose level, gender, BMI, drinking and smoking habit (P < 0.001). Diastolic blood pressure correlated also with a history of familial hypertension (P < 0.01). Due to the fact that age is correlated with blood glucose and blood pressure, we used age as a covariant in an analysis of covariance. After adjustment for age, the independent factors of DM and hypertension were determined. In the analysis of covariance, we found that BS was significantly related to the systolic blood pressure, BMI, diastolic blood pressure and a familial history of DM (P < O.OOl),as well as the level of education (P < 0.01). It was also shown that systolic blood pressure was significantly related to diastolic blood pressure, gender, BMI, drinking, smoking,
Gender Alcohol Smoking Diabetic FHa Hypertensive FHa Systolic BPa Diastolic BPa Glucose Age BMI Education
0.443* 0.587* -0.016 0.043** 0.156* 0.141* 0.021 0.035** 0.023 0.187*
aFH, family history; BP, blood pressure.
*P < 0001. . , **P < 0.01.
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
0.468* 0.001 0.011 0.131’ 0.126’ 0.026 0.082* 0.023 0.010 -0.007 0.013 0.141* 0.115’ 0.053* 0.1158 0.008 -0.012
0.128* [email protected]* -0.022 0.036** -0.083* 0.012 0.067*
0.009 0.040** -0.055* -0.109* -0.017 0.167:
0.792’ 0.188f 0.470* 0.123* -0.256*
0.144* 0.367* 0.138’ -0.211*
Matrix of correlation coefficient (Pearson) between risk factors of DM and hypertension with BS and blood pressure
0.228* 0.049* -0.172*
82 TABLE I
Stepwise multiple linear regression analysis predicting BS with different risk factors
Stepwise multiple linear regression analysis predicting diastolic blood pressure with different risk factors
8.084 7.036 4.575 -3.733 -2.768
Systolic BP BMI Hypertensive FH Drinking
91.32 4.903 3.965 2.547
Systolic BP Diabetic FH Education Hypertensive FH
a BP, blood pressure; FH, family history.
blood glucose level and a diabetic family history (P < 0.001). Diastolic blood pressure in addition was related to systolic blood pressure, gender, BMI, drinking, smoking, hypertensive family history and blood glucose level (P c 0.001). To study the different factors which contribute to the blood glucose and blood pressure levels, we used a stepwise multiple linear regression analysis of the data. In Table 7, the family history of DM, the family history of hypertension, the age, the education level and the systolic blood pressure significantly contributed to blood glucose level. Gender, diastolic blood pressure, age and blood glucose level statistically significantly contributed to systolic blood pressure (Table 8). A hypertensive family history, the systolic blood pressure and the BMI significantly contributed to diastolic blood pressure (Table 9). TABLE 8 Stepwise multiple linear regression analysis predicting systolic blood pressure with different risk factors. Variablea
Gender Diastolic BP
1.997 I.113 0.267 0.019 -1.119
Age BS Diabetic FH
a BP, blood pressure; FH, family history.
In our cross-section study, the mean systolic blood pressure decreased in slope of increment in the latter part of life. At the same time, we found of the mean diastolic blood pressure declined in the latter years as shown in Fig. 1. The result is compatible with a report from the United States in 1977 [lo]. The curve of the mean blood glucose level in the study (Fig. 2) is almost identical to the diastolic blood pressure. This was in part due to the increased mortality rate of hypertensive and diabetic patients in old age which made the population study underrepresented in the age group over 70 years. It is well known that BMI is an important factor in DM and hypertension [ 11,121. But from the curves of the mean blood pressure, the mean BS and the BMI, we thought that the BMI cannot explain the whole tendency for blood pressure and BS changes. The distinction between type 1 and type 2 DM was not made in this study. But for the Taiwanese population as far as the age onset of DM is concerned, only 0.7-0.8% of the total diabetic population had diabetes before the age of 15 . This rate is quite low when compared with other western nations. The data of an epidemiologic study of young-onset insulin dependent DM in Japan (0.54 yearly incidence/l00 000) and the United States (12.9 yearly incidence/100 000) indicate the low prevalence of type 1 DM in oriental countries [ 131. The diagnostic criteria, methodology used and the survey time are different, which make comparisons difficult. But the prevalence rate of DM in the
study is higher than most studies in Chinese populations reported before [ 14- 161. If the data is compared with those of the Shanghai survey (1.012%) or the Beijing survey (0.73%) , the prevalence rate of DM in this study is about two to three times higher than in mainland China. There are many reasons to interpret this discrepancy of diabetic prevalence rate between both Chinese population groups. The major reason to be taken into account is westernized life style and the dietary habit changes in the Taiwanese population. The data from Goto’s study done in 1962 and in 1973 found that the diabetic prevalence rate increased 5-fold [ 171. The reason for this increment is similar to the increase found in our study when compared with mainland China. Recent epidemiological study on DM in Pu-Li located at the center of Taiwan presented much higher diabetic prevalence rate (12.4%) than our result for those over 30 years old [ 181. For the same age group only 5.4% was diagnosed as DM in our study. Although the method and diagnostic criteria were different between two studies, but the community based and the time of survey almost the same. For those over 40 years old, the prevalence rate of DM in the present study is 8.0%. The prevalence rate of those patients between 40 and 68 years old is higher compared to the 1985 prevalence rate of 6.0% in Taipei . From an earlier report from Taiwan, the prevalence rate of DM in 1970 was only 5.05% . Our result is quite close to data published by Tai et al. (7.09%) which survey included Taipei city and 5 villages of Taiwan . In the present study, 57.2% of diabetic subjects knew they had DM before the survey. The prevalence data of undiagnosed DM - 42.8% in this district is thus quite close to the data reported in the United States (48.5%)  and Taiwan (38.9%) . In contrast with the DM prevalence rates, if we compare the prevalence rates for hypertension in this study with the 1963 and 1974 hypertensive survey carried out in Taiwan , we observe that there is not much change in the prevalence rate of hypertension and in the mean systolic and diastolic blood pressures. It is well known that hypertension is commonly found in association with DM [6,7,21]. The presence of both diseases may increase the risk of
cardiovascular complications  and cataract development . The data from the WHO states that in about 22-48.9% of diabetic patients hypertension was found . In our study, 24.4% of diabetic subjects had hypertension which was higher than in the normoglycemic subjects (5.2%). This association of DM and hypertension is confirmed by correlation coefficient and analysis of covariance studies. Our result consistent with the finding of Tai et al. that there seemed to be a tight link between hypertension and type 2 DM . The data is insufficient to establish a relationship between the educational level of the subjects and the prevalence rates of DM and hypertension. Reports from Israel [25,26] found that the educational level of the population was an important factor in the incidence rate of DM in adult men. These results were similar to the finding in our study - a higher educational level and lower diabetic prevalence rate. Our data indicated that after the adjustment for age in the analysis of covariance, educational level still significantly and negatively correlated with diabetic prevalence rate. In the district, a higher educational level promotes knowledge and emphasizes the importance to prevent chronic disease and to maintain an ideal body weight. In the matrix of correlation coefficient in our study (Table 6), the educational level correlated significantly and negatively with the BMI. In contrast to our study, a report from India concluded that the educational status was not correlated to prevalence of DM . Due to different races and cultures involved and different societies, the study may yield different results. A recent report regarding type 2 DM in Mexican Americans found that educational attainment was an important factor in the incidence of DM . As in other reports, the male/female ratio of DM ranged from 0.6 to 2.1 [29,30]. The ratio in our study was 0.75. But in correlation coefficient, we did not find that gender influenced blood glucose levels. In contrast to blood glucose levels, in our study gender influences the prevalence of hypertension. In analysis of covariance, both systolic blood pressure and diastolic blood pressure statistically correlated with gender (P < 0.001). From the Framingham study, it is known that males under the age of 60 years had a high inci-
dence of hypertension but females over 60 years old had a higher incidence of hypertension [3 11. A report from Japan indicated that in all the different age groups from 30 to over 70 years, the male subjects had a higher prevalence rate of hypertension than the females . In a longitudinal follow-up study in Sweden for the analysis factors of type 2 DM, it was found that smoking and alcohol consumption were not significantly related with the diabetic status . Similar results were observed in our study. In contrast, smoking and drinking in the analysis of covariance were significantly related to the systolic and diastolic blood pressure. Several reports from various countries reported a correlation between alcohol consumption and elevated blood pressure [34,35]. The relationship between smoking and blood pressure remains a controversial issue. Generally speaking, nicotine can acutely elevate systolic and diastolic blood pressure , but with regard to long-term smoking habits more information is needed to come to definitive conclusions. A family history of hypertension negatively correlated with the blood glucose level in this study. There are two reasons to explain this result. Firstly, if one family member has hypertension, other members may be more knowledgeable about the of preventing chronic diseases. importance Secondly, the premature death of diabetic patients with a history of familial hypertension were for obvious reasons not included in this survey. This may influence the results, in such a way that the history of familial hypertension correlated negatively with blood glucose levels. In conclusion, this study illustrates that the prevalence of DM in the studied district was 2.6% and for hypertension was 6.4%. For the age groups of 40 and older, the prevalence rates were 8.0% and 19.6%, respectively. From the data of the survey, we found that the most important factors contributing to the blood glucose levels were a family history of diabetes, a family history of hypertension, age and the systolic blood pressure. In contrast with its effect on blood glucose levels, gender, diastolic blood pressure, age and blood glucose were factors contributing to the systolic blood pressure. A family history of hypertension, systolic blood pressure and the BMI contributed to the diastolic blood pressure.
Acknowledgement This study was supported by grant from the Department of Health, Executive Yuan, Republic of China and Chang Gung Memorial Hospital. The authors are indebted to Professor Wen-Harn Pan who clearly directed the study and to Miss LihWey Lee for her statistical assistance. The authors also wish to thank the staff of Chang Gung Memorial Hospital, without whose support this study could not have been conducted. References I
Hu, H.H., Chu, F.L. and Chiang, B.N. et al. (1989)Prevalence of stroke in Taiwan. Stroke 20, 858-863. Tseng, W.P. (1980) A survey on deaths of patients with stroke. J. Formosan. Med. Assoc. 79, 625-630. Tseng, W.P. (1980) Outcome of untreated hypertensive in an agricultural population, a 15-year follow-up study. J. Formosan. Med. Assoc. 79, 556-563. Wu, K.W. and Wei, T.H. (1986) The prevalence rate of diabetes mellitus in group aged 40-69 in Taipei city, 1985. J. Formosan. Med. Assoc. 85, 1069-1077. Tai, T.Y., Yang, C.L. and Chang, C.J. et al. (1987) Epidemiology of diabetes mellitus among adult in Taiwan, R.O.C. In: S. Vannasaeng, W. Nitiyanant and S. Chandraprasert (Eds.), Epidemiology of Diabetes Mellitus. Bangkok, Thailand, pp. 42-48. Fuller, J.H. (1985) Epidemiology of hypertension associated with diabetes mellitus. Hypertension 7 (Suppl. 2). 3-7. Tai. T.Y.. Chuang, L.M., Chen, C.J. and Lin, B.J. (1991) Link between hypertension and diabetes mellitus epidemiological study of Chinese adults in Taiwan. Diabetes Care 14, 1013-1020. Juang. J.H., Hsu. L.F., Lin, J.D.. Huang, B.Y ., Huang, M.J. and Huang, H.S. (1987) Comparison of glucose reflectance meters for self-monitoring of blood glucose in diabetics. J. Formosan. Med. Assoc. 86, 667-671. Tsai. S.H. (1976) Clinical feature of diabetes in Taiwan. In: S. Baba, Y. Goto and I. Fukui (Eds.), Diabetes Mellitus in Asia. Amsterdam-Oxford, pp. 69-73. Roberts, J. and Maurer, K. Blood pressure levels of persons 6-74 years, United States. 1971-1974. National Center for Health Statistics 11 1 I. Sprtka. J.M.. Bender, A.P. and Jagger, H.G. (1988) Prevalence of hypertension and associated risk factors among diabetic individuals the three-city study. Diabetes Care 11. 17-22. Sitthi-Amorn, C.. Chandraprasert, S.. Bunnag, S.C. and Plengvidhya. C.S. (1989) The prevalence and risk factors of hypertension in Klong Toey Slum and Klong Toey government houses. Int. J. Epidemiol. 18, 89-94. Zimmet, P. (1982) Type II (non-insulin-dependent) diabetes - an epidemiological overview. Diabetologia 22, 399-4 I 1.
14 Cheah, Yeo, P.P., Thai, AC. et al. (1985) Epidemiology of diabetes mellitus in Singapore: comparison with other Asian countries. Ann. Acad. Med. 14, 232-239. 15 Tai, T.Y., Yang, C.L., Chang, C.J. et al. (1987) Epidemiology of diabetes mellitus in Taiwan, R.O.C. comparison between urban and rural areas. In: S. Vannasaeng, W. Nitiyanant and S. Chandraprasert (Eds.), Epidemiology of Diabetes Mellitus. Bangkok, Thailand, pp. 49-53. Zhong, X.Z. (1982) Diabetic mellitus survey in China. Chinese Med. J. 95, 423-430. Goto, Y. (1987) Vascular complication in diabetes in Japan. Adv. Metab. Dis. 9, 167-200. Chou, P.C., Chen, H.H. and Hsiao, K.J. (1992) Community-based epidemiological study on diabetes in Pu-Li, Taiwan. Diabetes Care 15, 81-89. Tsai. S.H. (1970) Epidemiology of diabetes mellitus in Taiwan. In: T. Tsujii and M. Wada (Eds.), Diabetes Mellitus in Asia, 1970. Amsterdam, Excerpta Medica, pp. 64-67. 20 Harris, M.I., Hadden, W.C., Knowler, W.C. and Bennett, P.H. (1987) Prevalence of diabetes and impaired glucose tolerance and plasma glucose levels in U.S. population aged 20-74 yr. Diabetes 36, 523-534. 21 Minaker, K.L. (1987) Aging and diabetes mellitus as risk factors for vascular disease. Am. J. Med. 82 (Suppl. IB), 47-53. 22 Masarei J.R.L., Kiiveri, H.T. and Stanton, K. (1986) Risk factors for cardiovascular disease in a diabetic population. Pathology 18, 89-93. 23 Szmyd, L. and Schwartz, B. (1989) Association of systemic hypertension and diabetes mellitus with cataract extraction. Ophthalmology 96, 1248-1252. 24 World Health Organization (1982) Vascular disease in diabetics. Report of WHO Multinational Study of Vascular Disease in diabetics. Internal Document NCD/OND/79.4. Geneva, WHO, pp. l-93. 25 Medalie, J.H., Papier, C.M., Goldbourt, U. and Herman, J.B. (1975) Major factors in the development of diabetes mellitus in 10 000 men. Arch. Intern. Med. 135, 81 I-817. 26 Medalie, J.H., Papier, C., Herman, J.B. et al. (1974)
Diabetes melhtus among 10 000 adult men. Israel J. Med. Sci. 10, 681-697. Gupta, O.P., Dave, P.S., Gupta, P.S. et al. (1976) Aetiological factors in the prevalence of diabetes in urban and rural population in India. In: S. Baba, Y. Goto and I. Fukui (Eds.), Diabetes Mellitus in Asia. AmsterdamOxford, pp. 23-34. Haffner, S.M., Hazuda, H.P., Mitchell, B.D., Patterson, J.K. and Stern, M.P. (1991) Increased incidence of type diabetes mellitus in Mexican Americans. Diabetes Care 14, 102-108. Spiegelman, M. and Mark, H.H. (1946) Age and sex variations in the prevalence and onset of diabetes mellitus. Am. J. Pub. Health 36, 26-33. Barrett-Connor, E. (1980) Factors associated with the distribution of fasting plasma glucose in an adult community. Am. J. Epidemiol. 112, 518-523. Dannenberg, A.L., Garrison, R.J. and Kannel, W.B. (1988) Incidence of Hypertension in the Framingham study. Am. J. Pub. Health 78, 676-679. Ueda, K. and Omae, T. (1985) Control of hypertension in a Japanese community. In: C.J. Bulpitt (Ed.), Handbook of hypertension, Vol. 6: Epidemiology of hypertension. Elsevier Science Publishers B.V., pp. 412-423. Ohlson, L.O., Larsson, B. and Bjorntorp, P. et al. (1988) Risk factors for type 2 (non-insulin-dependent) diabetes mellitus. Thirteen and one-half years of follow-up of the participants in a study of Swedish men born in 1913. Diabetologia 3 1, 798-805. Clark, V.A., Chapman, J.M. and Coulson. A.H. (1967) Effects of various factors on systolic and diastolic blood pressure in the Los Angeles heart study. J. Chron. Dis. 20, 571-581. Myrhed, M. (1974) Blood pressure. In: Alcohol consumption in relation to factors associated with ischemic heart disease; a co-twin control study. Acta Med. Stand. Supp. 567, 40-46. Mathers J.A.L., Patterson, M.C. and Levy, R.L. (1949) Some effects on the circulation of smoking cigarettes with varying nicotine content. Am. Heart J. 37, 612-619.