Insulin, The Renin-Angiotensin-Aldosterone System and Blood Pressure

INSULIN, THE RENIN-ANGIOTENSIN-ALDOSTERONE SYSTEM AND BLOOD PRESSURE JAMES F. TAIT and SYLVIA A. S. TAIT* There have been many studies of the possible correlation of insulin levels , e.g., blood insulin concentrations after overnight fasting or at various times after glucose loading, versus blood arterial blood pressure or the prevalence of hypertension. Subjects of different races with both normal and abnormal glucose tolerance, including patients with non-insulin dependent diabetes mellitus (NIDDM) have been studied. Role of Insulin in Determining Blood Pressure Considering all such reported studies, irrespective of the race of the subjects or whether the data are examined by uni- or multivariate analysis, the correlation between insulin levels and blood pressure is inconsistent [1, 2]. Multivariate analysis usually examines the confounding effects of variation in age, sex, and adiposity on the correlation of insulin levels and blood pressure. However, it has been shown that multivariate analysis involving interdependent variables, such as adiposity (e.g., BMI) and insulin levels, to form conclusions about the correlation of insulin level and blood pressure is ofdoubtful validity [3] . Also, there may be species and race variation in the strength of this correlation [4] . If data from Caucasians only are treated by univariate statistical analysis, then there is a more consistent correlation ofinsulin levels and blood pressure [5-11]. Univariate correlation in lean normotensive adult Caucasians was significant (P 40 years old) [55]. In Hisayama (Kyushu, southern Japan), also an area of relatively low salt intake , Ohmura, et al., found a highly significant correlation of insulin levels and blood pressures [56]. The corresponding results in Tokyo are more variable, possibly indicating differences in the birthplaces of the patients [57-59]. In general, present limited relevant data from Japan are not inconsistent with the predictions of the evolutionary theory concerning the 252 James F. Tait and Sylvia A. S. Tait ¦ Insulin and Blood Pressure inverse correlation ofinsulin-blood pressure and past salt intake. However, their quantity is inadequate for a critical assessment, and values for insulin levels are usually not given. Japanese, with normal glucose tolerance, show a positive correlation between blood pressure and salt intake [52; 53, p. 580] . These Japanese are not usually hyperinsulinemic. Therefore, it is not known whether the variation in insulin-blood pressure activities, as indicated in Japanese with NIDDM, may also be relatively important in determining the BP of normal Japanese. Conclusions on Differences in Insulin—Blood Pressure Activities The limited valid data presently available are not inconsistent with the evolutionary theory, which indicates that the present insulin-blood pressure activities of recently modernized hunter-gatherers are influenced by the environment of their ancestors, such as the conditions which determined their salt balance. However, until present insulin-blood pressure activities are determined by valid methodology and there is appropriate knowledge of the past environment, the theory cannot be rigorously tested. Nevertheless, present indications are that such investigations would be rewarding . Appendix: Correlation ofInsulin Level and Arterial Blood Pressure The results on die correlations of arterial blood pressure (mean [mBP] , systolic [sBP] or dBP [dBP] ) ? insulin blood levels (fasting [FI] and after glucose loading [RI] ) are from Meehan, et al., and also Zavaroni, et al., Pontremoli, et al., Berntorp, et al., Mykkanen, et al, Johnson, et al., and Schroll, et al. [5-11]. There were a total of 118 correlation studies from 42 reports (Table 1). The studies were divided according to the type of adult (>18 years), mainly Caucasian , subjects involved: (a) non-NIDDM, lean, normotensive; (b) non-NIDDM, obese, normotensive; and (c) early NIDDM. The studies were further divided as involving entirely or mainly males or females. However, in six studies there were equal numbers of male or females [60-62] . Other divisions were according to geographical situation (5 different European countries [63]) and age (18-, 21-, and 24year -olds [64] ) . To avoid bias, all divisions were initially maintained in the analysis. However, the results for FI and RI were pooled because the RI measurements were much smaller in number (about 25 percent) , and there was no difference statistically between the frequency of significant correlations for FI vs. BP compared with RI vs. BP. RESULTS The majority of studies (94 studies from 31 reports) were...