Why can high insulin levels indicate a risk for coronary heart disease?

Diabetologia, Sep 1994

Dr. A. Fontbonne

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


Why can high insulin levels indicate a risk for coronary heart disease?

Diabetologia Why can high insulin levels indicate a risk for coronary heart disease? A. Fontbonne INSERM Unit Villejuif France - 9 Springer-Verlag 1994 E v e r since it was first published that high circulating insulin levels were associated with an increased risk for coronary heart disease C H D in m e n [ 1-3 ], there has b e e n controversy as to how to interpret the finding. Indeed, times are long past w h e n the epidemiological identification of a "risk factor" automatically entailed suspicion as to its causal role towards the disease with which it was associated, but even so, the fact that any suspicion could fall on the h o r m o n e that "cures" diabetes s e e m e d intolerable to many. RJ Jarrett was and still is among the ardent defendants of insulin. In this issue of the journal, Diabetologia publishes his updated views on the topic of insulin and atherogenesis, and thankfully opens the opportunity for debate. Insulin, epidemiology, pathophysiology and common sense O n e can say there are two (complementary) ways to ascertain the plausibility of a causal relation between a "risk factor" and a disease: finding arguments within, and finding arguments outside the epidemiological field. Epidemiology has pointed to insulin as an ind e p e n d e n t and weakly but statistically significant prospective correlate of C H D in three cohorts of middle-aged m e n [ 1-3 ]. The fact that the correlate (fasting insulin or stimulated insulin) could vary between [ 1, 3 ] and even within studies [ 3, 4 ] most probAbbreviations: CHD, Coronary heart disease; NIDDM, noninsulin dependent diabetes mellitus. ably constitutes a r a n d o m finding, due to variability artifacts that are known to occur in multivariate analyses, rather than an inconsistency, and in any case does not alter the interpretation given the high correlation coefficients between the two measures in a healthy population. The arguments against causality n e e d not therefore be sought through a detailed critical review of the three above-mentioned studies. It is sufficient to r e m a r k that the result has been found so far only in healthy male middle-aged cohorts. That the correlation between plasma insulin levels and C H D was not evidenced in the Pima Indians [ 5 ] is no wonder: this population group experience so little cardiovascular disease as to render the search for subtle risk factors extremely difficult, if not meaningless. Far m o r e interesting is the fact that the association has never been proven in w o m e n [ 2, 6 ], nor seemingly in elderly m e n [7]. Inconsistent results in cohorts of diabetic or glucose-intolerant subjects also plead in favour of an absence of a causal relationship between hyperinsulinaemia and C H D [ 8 ]. If an excess of circulating insulin was actually toxic for the arteries, it would be difficult (though not impossible - specific protective factors in the female? Survival biases in the elderly?) to explain why this toxicity be expressed only in middle-aged non-diabetic men. Hence, epidemiology itself provides arguments against a causal relationship between insulin and cardiovascular disease. F r o m outside the field of epidemiology come arguments of another kind. Few biological studies have demonstrated a direct deleterious effect of insulin or insulin-like factors on the arterial wall - essentially a proliferation of muscle cells in the media, generally observed with insulin perfusions far above the physiological range [ 9 ]. O n the other hand several clinical studies have shown, and this is acknowledged by Jarrett himself as "'ample evidence", that hyperinsulinaemia is generally associated with a cluster of "well-documented risk factors for C H D " , among which are elevated blood pressure, low HDL-cholesterol levels and high triglycerides. This raises the possibility that hyperinsulinaemia appears to be linked to an increased cardiovascular risk because it indicates the presence of this cluster of plausibly atherogenic abnormalities. The question now is: why only in male middle-aged non-diabetic Caucasians? Jarrett proposes a somewhat surprising answer to this question, by assigning to hyperinsulinaemia a protective role against the deleterious effects of the "well-documented risk factors" to which it is associated. It is always difficult to derive pathophysiological explanations directly from epidemiological correlations, as we have just seen, but this hypothesis certainly challenges one's c o m m o n sense. If it were to be accepted, an alternative explanation would be n e e d e d to account for the negative correlation of HDL-cholesterol (a protective factor which increases - epidemiologically speaking - with total cholesterol) with cardiovascular events. More reasonably, one could suggest that high levels of circulating insulin m a y not always be the best indicator of the above-mentioned cluster of risk factors. For instance, it is well known that hyperinsulinaemia and obesity go hand in hand, which would m e a n rather "android" (centrally-distributed) obesity in men, and "gynoid" (peripherally-distributed) obesity in women. Jarrett is right to point out that an excess of visceral fat, typical of android obesity, is indeed associated with a greater risk of cardiovascular events, both in m e n and w o m e n [ 10 ], and it is actually together with android obesity that the cluster of cardiovascular risk factors can most easily be found [ 11 ]. Since in epidemiological cohorts hyperinsulinaemic w o m e n are generally gynoid obese and therefore not excessively at risk for cardiovascular disease, plasma insulin levels cannot be expected to relate to CHD. In N I D D M subjects - who are typically android obese -, hyperinsulinaemia is generally present, together with the accompanying cluster of risk factors. However, as the duration of the glycaemic disorder increases, the levels of plasma insulin decrease [ 12 ], and this hardly qualifies them, from a statistical point of view, for a positive correlation with cardiovascular events. Indeed, the best indicator of the cluster of C H D risk factors in this case appears to be a m o r e markedly android distribution of fat and a higher level of plasma triglycerides, which are the strongest predictors of C H D mortality in diabetic and glucose-intolerant subjects [ 13 ]. All this m a y readily account for what Jarrett calls '"inconsistencies" in the relationship between insulin and cardiovascular disease. Plasma insulin is not the most ubiquitous indicator of the existence of a potentially atherogenic syndrome. As its m e a s u r e m e n t is relatively time-consuming, expensive and poorly specific, a suggestion could be to replace it with a much cheaper and easier one: the m e a s u r e m e n t of the waist and the hip circumferences, the ratio of which gives a pretty good idea of the distribution of body fat, Insulin, not insulin, and the "small baby" hypothesis In the last paragraph of his article, Jarrett briefly mentions two recent British contributions m o r e or less related to the topic of insulin and cardiovascular disease. The first one concerns the imprecisions of the radioimmunological assay of insulin, the one commonly used in epidemiological studies, which does not distinguish between "true" insulin and its much less active precursors [ 14 ]. This m a y provide an explanation to the fact that "hyperinsulinaemic" m e n appear insulin resistant (they m a y have a lesser proportion of active insulin compared to "normoinsulinaemic" men). However, it is to be r e m e m b e r e d that in diabetic or glucose-intolerant subjects, supposedly the most deficient in active insulin despite usually "hyperinsulinaemic", plasma "total insulin-like molecules" are seldom if ever associated with cardiovascular complications, as we and Jarrett himself, in the first part of his paper, have mentioned. The hypothesis of Hales and Barker [ 15 ] of a fetal programming of future disease in adulthood is also m e n t i o n e d by Jarrett, though it seems peripheral to the present issue (unless Jarrett would now accuse insulin or its precursors to be risk determinants of cardiovascular complications?). It is also extremely questionable, since Hales and Barker's observations were m a d e on groups of elderly subjects, whose childhood took place immediately before or during the Second World War. Could it simply be that "small babies" with little or no capacity for storing fat did not survive the harsh circumstances of their childhood? H e n c e "small babies" who survived (studied much later by Barker) are more prone to obesity, N I D D M and cardiovascular disease later in life? A n d that actually a cluster of risk factors would protect "'small babies"? Conclusion Finally, Jarrett's article, though it m a y appear to verge on the polemic, proves that a kind of consensus has e m e r g e d between researchers in the field. Epidemiology has led us to identify (some m a y say: to rediscover, but this is another debate) an "at-risk" situation, i. e. the cluster of potentially atherogenic abnormalities generally associated with android obesity. Causality in the relationship between this cluster and cardiovascular events is plausible, since for at least some of its elements - and the list is not final - there exist physiological observations which constitute stronger arguments for a deleterious effect on arteries than is the case for insulin [ 16 ]. Having put insulin on trial has A.Fontbonne: High insulin levels a risk for coronary heart disease? s e r v e d a d o u b l e aim: it has f o u n d n o e v i d e n c e o f guilt, a n d it has r e v e a l e d t h a t h i g h levels o f t h e h o r m o n e c a n n e v e r t h e l e s s i n d i c a t e a s i t u a t i o n of g r e a t e r c a r d i o v a s c u l a r risk, u s u a l l y a s s o c i a t e d w i t h a n a n d r o i d distribut i o n of b o d y fat. 1. Py6r/~l/~ K ( 1979 ) Relationship of glucose tolerance and plasma insulin to the incidence of coronary heart disease: results from two population studies in Finland . Diabetes Care 2 : 131 - 141 2. Welborn TA , Wearne K ( 1979 ) Coronary heart disease incidence and cardiovascular mortality in Busselton with reference to glucose and insulin concentrations . Diabetes Care 2 : 154 - 160 3. Ducimetibre P , Eschw~ge E, Papoz L , Richard JL , Claude JR , Rosselin G ( 1980 ) Relationship of plasma insulin levels to the incidence of myocardial infarction and coronary heart disease mortality in a middle-aged population . Diabetologia 19 : 205 - 210 4. Fontbonne A , Charles MA , Thibult N e t al. ( 1991 ) Hyperinsulinaemia as a predictor of coronary heart disease mortality in a healthy population: The Paris Prospective Study, 15-year follow-up . Diabetologia 34 : 356 - 361 5. Liu QZ , Knowler WC , Nelson R G et al. ( 1992 ) Insulin treatment, endogenous insulin concentration and ECG abnormalities in diabetic Pima Indians: cross-sectional and prospective analyses . Diabetes 41 : 1141 - 1150 6. Fontbonne A ( 1991 ) Insulin, a sex hormone for cardiovascular risk ? Circulation 84 : 1442 - 1444 7. Welin L , Eriksson H , Larsson B , Ohlson LO , Sv~irdsudd K , Tibblin G ( 1992 ) Hyperinsulinaemia is not a major coronary risk factor in elderly men: the study of men born in 1913 . Diabetologia 35 : 766 - 770 8. Fontbonne A ( 1992 ) Relationship between diabetic dyslipoproteinaemia and coronary heart disease risk in subjects with non-insulin-dependent diabetes mellitus . Diabetes Metab Rev 7 : 179 - 189 9. Stout RW ( 1990 ) Insulin and atheroma . 20-yr perspective. Diabetes Care 13 : 631 - 654 10. Larsson B , Bengtsson C , Bj6rntorp P e t al. ( 1992 ) Is abdominal body fat distribution a major explanation for the sex difference in the incidence of myocardial infarction? The study of men born in 1913 and the study of women, G6teborg, Sweden . Am J Epidemio1135: 266 - 273 11. Kissebah A H ( 1991 ) Insulin resistance in visceral obesity . Int J Obesity 15 : 109 - 115 12. DeFronzo R A ( 1988 ) The triumvirate:/3-cell, muscle, liver. A collusion responsible for NIDDM . Diabetes 6 : 667 - 687 13. Fontbonne A , Thibult N , Eschw6ge E , Ducimeti6re P ( 1992 ) Body fat distribution and coronary heart disease mortality in subjects with impaired glucose tolerance or diabetes mellitus: The Paris Prospective Study, 15-year follow-up . Diabetologia 35 : 464 - 468 14. Temple RC , Carrington CA , Luzio SD et al. ( 1989 ) Insulin deficiency in non-insulin-dependent diabetes . Lancet I: 293 - 295 15. Hales CN , Barker DJP ( 1992 ) Type 2 (non-insulin-dependent) diabetes mellitus: The thrifty genotype hypothesis . Diabetologia 35 : 595 - 601 16. Austin MA ( 1991 ) Plasma triglyceride and coronary heart disease . Arterioscler Thromb 11 : 2 - 14

This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2FBF00400954.pdf

Dr. A. Fontbonne. Why can high insulin levels indicate a risk for coronary heart disease?, Diabetologia, 1994, 953-955, DOI: 10.1007/BF00400954