HomeLarge Type Edition
HOME ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
PubMed
Right arrow PubMed Citation
The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 56:M236-M242 (2001)
© 2001 The Gerontological Society of America

High Plasma Insulin and Lipids Profile in Older Individuals

The Italian Longitudinal Study on Aging

Stefania Maggia, Nadia Minicucia, Tamara Harrisb, Luciano Mottac, Marzia Baldereschid, Antonio Di Carlod, Domenico Inzitarid and Gaetano Crepaldi, for the ILSA Groupa

a National Research Council Center on Aging, University of Padua, Italy
b Epidemiology, Demography, and Biometry Program, National Institute on Aging, National Institutes of Health, Bethesda, Maryland.
c University of Catania, Italy
d National Research Council of Florence, Italy.

Stefania Maggi, c/o Clinica Medica 1°, University of Padua, Via Giustiniani, 2, 35128 PADOVA, Italy E-mail: smaggi{at}ux1.unipd.it.

Decision Editor: John E. Morley, MB, BCh


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 References
 
Background. The inverse relationship of insulin level to high-density lipoprotein (HDL)-cholesterol and its positive association with hypertriglyceridemia has been demonstrated in several studies; however, the relationship of insulin to low-density lipoprotein (LDL)-cholesterol in elderly persons is not clear. This study investigates the relationships of fasting plasma insulin and selected metabolic and biological risk factors in an aged population.

Methods. The present study is based on a cross-sectional analysis of the data collected at baseline of the Italian Longitudinal Study on Aging in 1992 on a random sample of 5632 Italians aged 65–84 years. Analyses were performed to compare the distribution of risk factors, such as blood level of lipids, creatinine, albumin, fibrinogen, apolipoprotein A-1 and B, blood pressure, and body mass index (BMI), by quartiles of insulin, in both diabetic and nondiabetic participants.

Results. Significantly higher levels of triglycerides and BMI and lower levels of HDL-cholesterol were found in the upper quartile of insulin among nondiabetic individuals. In men, we also found significantly higher levels of systolic and diastolic blood pressure. The same trend for these variables, although not significant for HDL-cholesterol and blood pressure, was seen in diabetic men. In diabetic women, total and LDL-cholesterol were significantly lower in the highest insulin quartile (p < .001), while no significant differences were seen in nondiabetic women or in men. We also found higher levels of white blood cells in the highest insulin quartile of diabetic women.

Conclusions. These results, apparently in disagreement with earlier reports on the clustering of cardiovascular disease risk factors in hyperinsulinemic individuals, could be due to the high frequency of chronic inflammation and the high prevalence of urinary infections in older diabetic women.

HYPERINSULINEMIA, an indicator of insulin resistance, is generally considered a risk factor for cardiovascular disease, but the physiological mechanism for this association remains unclear (1)(2)(3)(4). It is still contested, however, whether hyperinsulinemia is an independent risk factor for coronary artery diseases in elderly persons (5)(6)(7). Studies on older individuals are still scanty, and recent data on a cohort of elderly hypertensive patients with central obesity and impaired glucose control did not show any association with fasting hyperinsulinemia, suggesting that syndrome X did not occur in this cohort (8). Moreover, the relationship of insulin to low-density lipoprotein (LDL)-cholesterol (CH) is not yet clear. A large epidemiologic study reported lower levels of LDL-CH in diabetic individuals (9), and another cross-sectional study (10) found an inverse association of hyperinsulinemia with LDL-CH in an older population after adjustment for age, BMI, blood glucose, and serum triglycerides. Such an association has been confirmed in a follow-up of the same cohort that excluded the effect of a selective mortality of individuals with high LDL-CH and hyperinsulinemia (11).

We analyzed the relationship of fasting plasma insulin and selected metabolic risk factors, such as blood levels of lipids, creatinine, albumin, fibrinogen, apolipoprotein A-1 and B (apo A-1 and apo B), blood pressure, and body mass index (BMI), in older individuals.


   Methods
Top
 Abstract
 Methods
Results
 Discussion
 References
 
The Italian Longitudinal Study on Aging (ILSA) has been described in detail elsewhere (12). A random sample of 5632 individuals aged 65–84 years, including community-dwelling and institutionalized persons, stratified by age and sex using an equal allocation strategy, was identified on the demographic lists of the registry office of eight municipalities. Eighty-eight subjects of each gender in four age groups (65–69, 70–74, 75–79, and 80–84 years) were included in the study sample. The ILSA design includes cross-sectional and longitudinal components. This paper is based on data from the cross-sectional study. The baseline survey had two phases: (i) a first phase, administered to all participants, included a personal interview on self-reported conditions, laboratory tests based on a blood sample obtained after an overnight fast, a physician examination, and selected diagnostic tests (e.g., spirometry, electrocardiography, and retinography); (ii) a second phase, administered to selected participants, consisted of the clinical confirmation of suspected cases. The clinical confirmation was made by a specialist (internist or neurologist) through a visit and a review of medical records.

The data presented in this report were obtained during the baseline examination carried out between March 1992 and June 1993.

Diagnosis of Diabetes
The diagnostic criteria used to define the prevalence rates of the investigated conditions included a screening phase and a clinical confirmation of the disease. Details on the diagnostic criteria and data on the prevalence rates of diseases have been previously published (13). In particular, for diabetes, the criteria for the positive screening were (i) self-reported diagnosis or (ii) fasting glucose plasma >=140 mg/dl. The self-reported information was based on the questions, "Has a doctor ever told you that you are affected by diabetes?" and "Are you currently taking insulin or anti-diabetic treatments?" Those who reported the condition independently of the treatment and those who did not report the condition but had glycemia >=140mg/dl and were confirmed positive by the ILSA internist (after a second glucose determination >=140 mg/dl) were included as prevalent cases. For those who screened positively but did not undergo the clinical examination with the ILSA internist, the diagnosis was confirmed by their general practitioners (GPs).

Blood Determinations
Plasma lipids (total and HDL-cholesterol and triglycerides) and glucose were measured using standard enzymatic methods: insulin and apo A-1 and apo B were measured by radioimmunoassay; fibrinogen, glycohemoglobin (HbA1C), and proteins were measured by electrophoresis; and the cell blood counts were performed by an automatic counter. The lipids and insulin determinations were centrally analyzed at the Catania University's laboratory. LDL-cholesterol was calculated by Friedewald's equation (14).

Statistical Analyses
All statistical analyses were performed using the SAS statistical analysis package (SAS Institute Inc, Cary, NC) (15). Quartiles of insulin were based on the total sample distribution (n = 2768), and variables in insulin quartiles were compared using the General Linear Model procedure. A t test was performed to identify the significant differences among the means. Chi-square was used to test the association between categorical variables and the trend for age.


   Results
Top
 Abstract
 Methods
 Results
Discussion
 References
 
Of the 5632 subjects originally sampled, 170 were not eligible at the beginning of the survey because they died or moved before the prevalence day. In the interview phase there were 953 nonrespondents (17.4%), and among the 4509 screened individuals, 21.7% screened positive for the condition (n = 979). In phase two, the suspected cases were invited for a clinical examination, and 757 were visited by an internist, while information for 199 subjects was obtained by the GPs. Sixty percent (589) were diagnosed as diabetics, 364 (37%) were nondiabetic, and 26 subjects were not examined and were therefore excluded from this analysis. Hence, the total sample was comprised of 589 diabetic and 3894 nondiabetic subjects.

The overall prevalence of diabetes was about 13% in both men and women, and no significant trend with age was detected. Table 1 shows the distribution of participants by sex and diabetes status in the quartiles of insulin. About 40% of the participants refused to have a blood sample taken; therefore, the analysis by quartile of insulin is based on a sample size of 372 diabetic and 2396 nondiabetic subjects. About 16% of men and women with diabetes are in the lowest quartile, versus 27% of those without the disease. Diabetics had consistently higher levels of insulin: 42.3% of men and 36.9% of women with diabetes were in the highest quartile versus 23.2% of men and 22.9% of women without the disease. A significant association between diabetic status and quartile of insulin (p < .0001) was found in both sexes.


View this table:
[in this window]
[in a new window]
  		Table 1. Distribution of Diabetic and Nondiabetic Participants by Quartile of Insulin and Sex

 
Selected characteristics by quartile of insulin are presented in Table 2 and Table 3 . Among nondiabetic individuals, significantly higher levels of glucose, triglycerides, BMI, and HbA1c and lower levels of high-density lipoprotein cholesterol (HDL-CH) and apo A-1 are found in the upper quartile contrasted with the first quartile in both men and women. Higher insulin levels were associated with higher values of creatinine, albumin, fibrinogen, apolipoprotein B, and systolic and diastolic blood pressure in men. In women only, higher levels of white blood cells were associated with higher insulin. However, there was no association of LDL-CH with hyperinsulinemia in either nondiabetic men or women. Fewer associations were seen among diabetics (Table 4 and Table 5 ), with significantly higher levels of triglycerides and BMI and lower levels of apo A-1 found in both sexes in relation to increasing insulin. Higher levels of glucose were demonstrated for men and higher levels of white blood cells for women. In men, the upper quartile of insulin had lower mean values for LDL-CH, but this mean value did not differ significantly from the mean of the lowest quartile of insulin. In women, differences in means between the highest and the lowest quartile for total cholesterol and LDL-CH were statistically significant, with lower values in the highest quartile.


View this table:
[in this window]
[in a new window]
  		Table 2. Blood Glucose, Serum Lipids, and Clinical Characteristics by Quartile of Plasma Insulin (mU/ml) in the Male Nondiabetic Population, Adjusted by Age

 

View this table:
[in this window]
[in a new window]
  		Table 3. Blood Glucose, Serum Lipids, and Clinical Characteristics by Quartile of Plasma Insulin (mU/ml) in the Female Nondiabetic Population, Adjusted by Age

 

View this table:
[in this window]
[in a new window]
  		Table 4. Blood Glucose, Serum Lipids, and Clinical Characteristics by Quartile of Plasma Insulin (mU/ml) in the Male Diabetic Population, Adjusted by Age

 

View this table:
[in this window]
[in a new window]
  		Table 5. Blood Glucose, Serum Lipids, and Clinical Characteristics by Quartile of Plasma Insulin (mU/ml) in the Female Diabetic Population, Adjusted by Age

 
We performed a bivariate analysis, which included all the biological factors under study, to find out if any other factors could explain the association of low LDL and insulin level. The association of low LDL and hyperinsulinemia was not significant in nondiabetic men and women or in diabetic men after adjusting for age and triglycerides, fibrinogen, BMI, glycemia, creatinine, albumin, white blood cells, apo A-1 and apo B, blood pressure, and HbA1c. However, as presented in Table 6 , such an association was significant in diabetic women after all adjustments.


View this table:
[in this window]
[in a new window]
  		Table 6. Low Density Lipoprotein Cholesterol by Quartile of Plasma Insulin (mU/ml) in the Female Diabetic Population

 

   Discussion
Top
 Abstract
 Methods
 Results
 Discussion
References
 
This population-based study shows an inverse association between plasma insulin and total cholesterol, mainly due to the lower level of LDL cholesterol in older diabetic women. In nondiabetic participants we found the expected CVD risk profile in the upper quartile of insulin, with higher values of the BMI and of the levels of triglycerides, while there were lower levels of HDL-CH and apo A-1. However, among diabetic participants, we did not observe any differences in the levels of HDL-CH across quartiles, and in women we found significantly lower levels of total cholesterol and LDL-CH, with higher levels of white blood cells in the upper insulin quartile. This association held true after adjusting for all potential confounding factors, such as age, triglycerides, fibrinogen, BMI, glucose, and Apo B. Strandberg and colleagues (10) suggested that high plasma insulin is associated with lower LDL-CH in elderly individuals, independently of their diabetic status and even after adjusting for age, gender, BMI, and serum triglycerides. The association was reported as an inverse U-shaped curve. However, data are not presented by gender, and the diagnosis of diabetes seems to be based on a single blood glucose determination. More than 60% of the diabetic subjects were women, and it might be hypothesized that their majority could explain the association of LDL-CH and hyperinsulinemia found in diabetics. However, the association found in nondiabetic individuals was not confirmed in our data. Our findings, confirming the results only in diabetic women, could have several explanations. First, there is evidence for inflammation as a cause of hypocholesterolemia in older people (16), and diabetic women are at a higher risk, for example, for urinary infections compared with diabetic men and with the general population. In our study, diabetic women with hyperinsulinemia had significantly higher levels of white blood cells, although such values (about 7.000/mm3) do not seem to be clinically significant for infective or inflammatory disorders. Other studies have found low CH levels in both diabetic men and women (9)(17).

Maggi and colleagues (17) hypothesized the effect of selective survival of diabetics with a better CVD risk profile, but why this selective mortality should affect only women and not men in the present study is not clear. Moreover, a prospective study by Lindberg and colleagues (11) does not support this hypothesis and found that the combination of high LDL-CH with high insulin was not associated with any increase in the crude mortality rate in their cohort.

Manolio and colleagues (9) suggested that the association between low cholesterol levels among diabetics could be due to the cholesterol-lowering effect of hypoglycemic drugs. However, in our sample the percent of treated men and women did not differ significantly. One potential explanation for the sex difference in the total and LDL-CH, therefore, could be that the compliance to the treatment is better among women than among men. We have not investigated this issue in the present study; however, if this were the explanation, we would have expected a higher HDL-CH level among diabetic women. CH absorption has been found to be lower in diabetic individuals with hypertriglyceridemia (18), but, if this were the explanation for our findings we would have seen lower CH level in both sexes, given that diabetic men and women have similar levels of triglycerides (TG) in the upper quartile of insulin. Other more recent studies (19)(20) show an inverse association between cholesterol absorption and insulin in diabetic and nondiabetic patients. Finally, it has been shown that hyperglycemia is associated with an increased LDL-CH turnover in kinetic studies (21). This would explain the association of lower LDL-CH in the upper quartile of insulin with increased catabolism, but again we are unable to explain the gender difference in such an association.

The LDL-CH concentration was estimated by Friedewald's equation. It has been reported that in diabetic patients it should not be used because of frequent hypertryglyceridemia and abnormal very low-density lipoprotein or LDL composition, which perturb the Friedewald's equation (22)(23). However, a more recent study (24) has established an excellent agreement between the direct LDL-CH assay and the Friedewald's formula in normotriglyceremic diabetic individuals. Given that our sample has concentrations <4.5 mmol/l, we could assume the reliability and accuracy of the formula. Moreover, we have ruled out the possibility that the lower LDL-CH levels in the highest quartile reflected a higher TG concentration, given that after adjusting for TG and age in a bivariate analysis, the association was still significant.

Although a positive association of low CH levels with mortality has been reported in several studies (25)(26), a causal relation has not been demonstrated. The follow-up of this cohort, with the assessment of the mortality rate, will help to clarify several issues presented in this study. First of all, it could be that low LDL-CH constitutes the alleged protective factor associated with hyperinsulinemia suggested by Jarrett (27). Several studies have failed to confirm an association between hyperinsulinemia and CVD events (5)(6)(7); if we confirm these findings in older diabetic women, we could probably support the protective effect of low LDL.

The insulin level in our population is higher than previously reported by Strandberg and colleagues (10), but, in spite of the fact that both studies used radio immuno assay methods, a large variability in the measure and a lack of comparability across studies is expected (28).

The present study has some methodological limitations. First of all, it is based on a cross-sectional assessment, and therefore a survival effect cannot be ruled out. Older women with diabetes and high LDL and insulin levels might experience a higher mortality rate, and therefore our population might be predominantly made up of selected survivors. We will test this hypothesis by means of the five-year follow-up of this cohort, available shortly. Moreover, although the rate of nonrespondents did not differ from previous studies in the elderly and was comparable to the participants in terms of age and gender distribution, a selection bias could have occurred, and therefore the inclusion of a healthier sample compared with the general older population cannot be excluded.

The ILSA study has several strengths. First of all, the diagnosis of diabetes was clinically ascertained, and most of the laboratory determinations were centralized to avoid intercenter variability. Criteria for the diagnosis of the diseases were standardized, and interrater reliability studies were carried out in all participating centers.

In conclusion, this study found lower LDL cholesterol levels among older women with diabetes and higher insulin levels. Longitudinal studies are needed to confirm these findings and to clarify the role of low LDL-cholesterol as a potential "predictor of adverse events" or, to the contrary, as a "protective factor" associated with hyperinsulinemia in the elderly.


   Acknowledgments
 
The ILSA was supported by the Italian Research Council, Progetto Finalizzato Invecchiamento, and by the Instituto Superiore di Sanit", Progetto Epidemiologia dell'anziano. This work was completed during Dr. Maggi's visit award at National Institutes of Health; National Institute on Aging; Epidemiology, Demographic, and Biometry Program, Bethesda, MD.


   Footnotes
 
Members of the (Italian Longitudinal Study on Aging) ILSA Working Group are: University of Milan: Guglielmo Scarlato, Livia Candelise, Elio Scarpini; University of Padua: Leontino Battistin, Giuliano Enzi, Francesco Grigoletto, Marta Bressan; CNR-LADSEB Padua: Giovanni Bortolan; CNR Genoa: Carlo Loeb; University of Genoa: Carlo Gandolfo; San Raffaele Institute, Milan: Nicola Canal, Massimo Franceschi; USSL 10, Florence: Augusto Ghetti, Rossano Vergassola; INRCA, Fermo (AP): Salvatore Bonaiuto; University of Naples: Franco Rengo; University of Bari: Antonio Capurso, Paolo Livrea; University of Catania: Luciano Motta.

Received April 19, 2000

Accepted May 8, 2000


   References
Top
 Abstract
 Methods
 Results
 Discussion
 References
 

  1. Després JP, Lamarche B, Mauriege P, et al. 1996. Hyperinsulinemia as an independent risk factor for ischemic heart disease. N Engl J Med 334:952-957. [Abstract/Free Full Text]
  2. Ferrannini E, Haffner SM, Mitchell BD, Stern MP, 1991. Hyperinsulinaemia: the key feature of a cardiovascular and metabolic syndrome. Diabetologia 34:416-422. [Medline]
  3. Pyorala 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 2:121-141.
  4. Ducimetiere P, Eschwege 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. [Medline]
  5. Ferrara A, Barrett-Connor EL, Edelstein SL, 1994. Hyperinsulinemia does not increase the risk of fatal cardiovascular disease in the elderly men or women without diabetes: the Rancho Bernardo Study, 1984–1991. Am J Epidemiol 140:857-869. [Abstract/Free Full Text]
  6. Welin L, Eriksson H, Larsson B, Ohlson LO, Svardsudd K, Tibblin G, 1992. Hyperinsulinemia is not a major coronary risk factor in elderly men: the study of men born in 1913. Diabetologia 35:766-770. [Medline]
  7. Welborn TA, Weane K, 1979. Coronary heart disease incidence and cardiovascular mortality in Busselton with reference to glucose and insulin concentrations. Diabetes Care 2:154-160. [Abstract]
  8. Johnson KC, Graney MJ, Applegate WB, Kitabchi AE, Runyan JW, Rutan GH, 1999. Does syndrome X exist in hypertensive elderly persons with impaired glycemic control?. J Gerontol Med Sci 54A:M571-M576. [Abstract]
  9. Manolio TA, Ettinger WH, Tracy RP, et al. 1993. for the CHS Collaborative Research Group. Epidemiology of low cholesterol levels in older adults: The Cardiovascular Health Study. Circulation 87:728-737. [Abstract/Free Full Text]
  10. Strandberg TE, Tilvis RS, Lindberg O, et al. 1996. High plasma insulin is associated with lower LDL cholesterol in elderly individuals. Atherosclerosis 121:267-273. [Medline]
  11. Lindberg O, Tilvis RS, Strandberg TE, Lehtonen A, Ehnholm C, Tuomilehto J, 1998. Inverse association of serum cholesterol with plasma insulin in the elderly: cross national and prospective analyses. Aging Clin Exp Res 10:137-140.
  12. Maggi S, Zucchetto M, Grigoletto F, et al. 1994. for the ILSA group. The Italian Longitudinal Study on Aging (ILSA): design and methods. Aging Clin Exp Res 6:464-473.
  13. The ILSA Group1997. Prevalence of chronic diseases in older italians: comparing self-reported and clinical diagnosis. Int J Epidemiol 26:995-1002. [Abstract/Free Full Text]
  14. Friedewald WT, Levy RI, Fredrickson DS, 1972. Estimation of the concentration of low density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem 18:499-502. [Abstract]
  15. SAS Institute Inc1989. SAS/STAT User's Guide, Version 6 4th ed. SAS Institute, Inc, Cary, NC.
  16. Ettinger WE, Harris T, Verdery RB, Tracy RP, Kouba E, 1995. Evidence for inflammation as a cause of hypocholesterolemia in older people. JAGS 43:264-266.
  17. Maggi S, Bush T, Hale WE, 1990. Diabetes mellitus and other cardiovascular risk factors in an elderly population. Age and Aging 19:173-178.
  18. Briones ER, Steiger DL, Palumbo PJ, et al. 1986. Sterol excretion and cholesterol absorption in diabetics and nondiabetics with and without hyperlipidemia. Am J Clin Nutr 44:353-361. [Abstract/Free Full Text]
  19. Gylling H, Miettinen TA, 1997. Cholesterol absorption, synthesis and LDL-metabolism in NIDDM. Diabetes Care 20:90-95. [Abstract]
  20. Strandberg TE, Salomaa V, Vanhanen H, Miettinen TA, 1996. Associations of fasting blood glucose with cholesterol absorption and synthesis in non-diabetic middle-aged men. Diabetes 45:755-761. [Abstract]
  21. Kissebah AH, Alfarsi S, Evans DJ, Adams PW, 1983. Plasma low density lipoprotein transport kinetics in noninsulin-dependent diabetes mellitus. J Clin Invest 71:655-667.
  22. Kasama T, Yoshino G, Iwatani I, et al. 1987. Increased cholesterol concentration in intermediate density lipoprotein fraction of normolipidemic non-insulin-dependent diabetics. Atherosclerosis 63:263-266. [Medline]
  23. Joven J, Vilella E, Costa B, Turner PR, Richart C, Masana L, 1989. Concentrations of lipids and apolipoproteins in patients with clinically well-controlled insulin-dependent and non-insulin-dependent diabetes. Clin Chem 35:813-816. [Abstract/Free Full Text]
  24. Whiting MJ, Shephard MD, Tallis GA, 1997. Measurement of plasma LDL cholesterol in patients with diabetes. Diabetes Care 20:12-14. [Abstract]
  25. Yano K, Reed DM, MacLean CJ, 1989. Serum cholesterol and hemorrhagic stroke in the Honolulu Heart Program. Stroke 20:1460-1465. [Abstract/Free Full Text]
  26. Muldoon MF, Manuck SB, Matthews KA, 1990. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. Br Med J 301:309-314.
  27. Jarrett RJ, 1994. Why is insulin not a risk factor for coronary heart disease?. Diabetologia 37:945-947. [Medline]
  28. Robbins DC, Andersen L, Bowsher R, et al. 1996. Report of the American Diabetes Association's Task Force on standardization of the insulin assay. Diabetes 45:242-256. [Abstract]



This article has been cited by other articles:


Home page
 Journals of Gerontology Series A: Biological Sciences and Medical SciencesHome page
W. S. Aronow
Guest Editorial: Should Hypercholesterolemia in Older Persons Be Treated to Reduce Cardiovascular Events?
J. Gerontol. A Biol. Sci. Med. Sci., July 1, 2002; 57(7): M411 - 413.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
PubMed
Right arrow PubMed Citation

HOME ARCHIVE SEARCH TABLE OF CONTENTS