Lipid risk factors among elderly with normal fasting glucose, impaired fasting glucose and type 2 diabetes mellitus. The Italian longitudinal study on aging

Nutrition, Metabolism & Cardiovascular Diseases (2013) 23, 220e226

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nmcd

Lipid risk factors among elderly with normal fasting glucose, impaired fasting glucose and type 2 diabetes mellitus. The Italian longitudinal study on aging M. Noale*,1, S. Maggi 1, S. Zanoni 1, F. Limongi 1, S. Zambon 1, G. Crepaldi 1 CNR-Institute of Neuroscience, Padova, Via Giustiniani, 2, 35128 Padova, Italy Received 21 March 2011; received in revised form 23 June 2011; accepted 23 June 2011 Available online 19 September 2011

KEYWORDS Lipid risk factors; Diabetes; Impaired fasting glucose; Normal fasting glucose; Elderly; Factor analysis

Abstract Background and aims: Three groups of subjects were identified within a representative sample of older Italians: subjects with normal fasting glucose (NFG), with impaired fasting glucose (IFG) or with type 2 diabetes mellitus (T2D). The aim of the present study was to evaluate the relationship among plasma lipids, lipoproteins, other metabolic factors in the three groups, and their role in predicting total fatal events. Methods and results: 2422 subjects, aged 65e84 years, taking part into the Italian Longitudinal Study on Aging were included in the analyses. Factor analysis was conducted separately for men and women. Factor scores were used as independent variables in Cox Proportional Hazard models, to determine factors predicting death at the follow-up in NFG, IFG and T2D subjects. Four major factors were found for men (“insulin resistance”, “body size”, “total cholesterol”, “HDL cholesterol”) and four also for women (“insulin resistance”, “total cholesterol”, “body size”, “HDL cholesterol”). For NFG and IFG men, and for both T2D men and women, the “HDL cholesterol” was a significant protective factor for total deaths (NFG men: HR Z 0.79, 95% CI 0.67e0.93; IFG men: HR Z 0.59, 95% CI 0.45e0.79; T2D men: HR Z 0.55, 95% CI 0.34 e0.89; T2D women: HR Z 0.61, 95% CI 0.44e0.86). Among NFG women, the “body size” factor was also a protective factor with respect to total deaths (HR Z 0.74, 95% CI 0.57e0.95). Conclusion: A factor including HDL Cholesterol and Apo A-I showed protection against all-cause mortality in older men, independently from the glycemia level, and in women only in those diagnosed with T2D. ª 2011 Elsevier B.V. All rights reserved.

Abbreviations: NFG, normal fasting glucose; IFG, impaired fasting glucose; T2D, type 2 diabetes mellitus; ILSA, Italian longitudinal study on aging; HDL-C, HDL cholesterol; HOMA IR, homeostasis model IR; IHD, ischemic heart disease; CVD, cardiovascular disease. * Corresponding author. Tel.: þ39 0498218899; fax: þ39 0498211818. E-mail address: [email protected] (M. Noale). 1 for the ILSA Working group. 0939-4753/$ - see front matter ª 2011 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.numecd.2011.06.004

Lipid risk factors among elderly with normal fasting glucose, impaired fasting glucose and type 2 diabetes mellitus

Introduction Numerous studies [1e6] and also a series of large metaanalyses [7e9] evidenced that type 2 diabetes mellitus (T2D) increases the risk of all-cause mortality and cardiovascular disease mortality. Moreover, some studies reported that among type-2 diabetic subjects, lipids rather than glycemic parameters were predictors of total mortality [10,11]. However, it is not completely clear which lipoprotein abnormality is predominant among diabetic elderly [12]. Subjects with diabetes are characterized by hypertriglyceridemia and low HDL cholesterol (HLD-C [13];). Less unanimous is discussion on LDL cholesterol, since some studies reported normal values [13], other studies higher values [14], other lower LDL cholesterol values [15], in diabetic subjects with respect to non diabetic controls. Moreover, specific lipid characteristics of NFG and IFG subjects are not well defined. Based on these considerations, the present analyses were developed to investigate relationships among plasma lipids, lipoproteins and other metabolic factors in a representative sample of older Italians, stratified into NFG, IFG, T2D subjects, and to evaluate their role in predicting total fatal events.

Methods The Italian Longitudinal Study on Aging (ILSA) has been described in detail elsewhere [16]. A random sample of 5632 individuals aged 65e84 years, including both communitydwelling 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. For each municipality, eighty-eight subjects of each gender in four age groups (65e69, 70e74, 75e79, 80e84 years) were included in the study sample. This paper is based on data from both the cross-sectional (1992) and the longitudinal wave of 1996 of the ILSA study. The baseline survey had two phases: a first phase, administered to all participants, included: a personal interview on socio-demographic characteristics, health habits, self-reported conditions, family and medical history, drugs use; a nurse examination, with blood pressure and pulse rate measurements, and a fasting blood sample evaluation; a physician examination. A second phase, administered to participants who screened positive to the first phase, consisted of clinical confirmation of suspected cases of cardiovascular diseases, diabetes, parkinsonism, stroke, dementia and peripheral neuropathy by a specialist (internist or neurologist) through a visit and the review of medical records. Diagnostic criteria and the relative prevalence rates have been published elsewhere [17]. The study protocol and the aims of the study were explained to each subject before they provided written informed consent. The protocol was approved by the Ethics Committees of the participating centers.

Blood determinations Plasma lipids (total cholesterol, HDL-C and triglycerides) and glucose were measured using standard enzymatic

221

methods; insulin, Apo A-I and Apo B were measured by radioimmunoassay; fibrinogen, glycohemoglobin (HbA1C) and proteins by electrophoresis; 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 [18]. Homeostasis model IR (HOMA IR) was calculated considering the formula [fasting insulin (mU/ml) * fasting plasma glucose (mmol/l)]/22.5.

Anthropometric measures Height and weight were measured with the subject barefooted and lightly dressed. Body weight was measured on a balance beam platform scale (Salus, Milan, Italy) to the nearest 0$1 kg. Height was taken by a stadiometer (Salus) at head level to the nearest centimeter with the subject standing barefoot, with feet together. Body Mass Index (BMI) was calculated as weight divided by height squared (kg/m2). Circumferences were measured to the nearest centimeter using a flexible steel tape, with the subject standing. The abdominal circumference (waist) was measured at the end of expiration, by wrapping the tape at the level of the umbilicus [19].

Diagnostic criteria for diabetes The criteria for positive screening for diabetes were: (1) self-reported diagnosis or treatment, (2) fasting glycemia 140 mg/dl (7.78 mmol/L). Those who screened positive were evaluated by an internist, after a second glucose determination 140 mg/dl and/or review of medical records. Prevalence rates of diabetes for the ILSA study were calculated in 1994, and therefore earlier diagnostic criteria were used (140 mg/dl) instead of the latest criteria (126 mg/dl [20];). As a consequence, a group of subjects with fasting glycemia between 126 mg/dl and 139 mg/dl were classified in the IFG group.

Study population Among the 5632 subjects originally sampled, 1096 (19.5%, weighted data) did non participate to the first phase of the baseline assessment. Non-respondents were found to be older (73.1  5.6 vs 72.5  5.6 years, p Z 0.0008) and with a higher percentage of women (64.2 vs 57.5, p < 0.0001). Among the 4536 for whom socio-demographic characteristics were available at the baseline, 2422 subjects had complete information on anthropometric measures, blood determinations and diabetes status; their mean age was 71.9  5.5 years, and 55% were females. Subjects were then categorized into three subgroups, according to their baseline fasting glucose level: (1) subjects with NFG (glucose <100 mg/dl, n Z 1423); (2) subjects with IFG (glucose 100 mg/dl but not with T2D, n Z 697); (3) subjects with T2D (n Z 302).

222

Statistical analysis To generalize the ILSA sample to the Italian population and to take into account the sample design, a set of weights was defined according to the age distribution of the Italian reference population (1991 Census) and the sample fraction, and applied into the analysis. Associations between groups of subjects defined and demographic and clinical characteristics were investigated through the ManteleHaenszel c2 test for trend for categorical variables, and ordered JonckheereeTerpstra test for continuous variables. Exploratory factor analysis was used to identify specific clusters from the lipid profiles of subjects, on the basis of the correlation between measures, stratifying for sex and study groups (NFG, IFG, T2D). The measurements included in the factor analyses were plasma triglycerides, total cholesterol, HDL-C, Apo A-I, Apo B, fasting insulin, HOMA IR, fasting plasma glucose, BMI, abdominal circumference. Factor extraction was conducted using the method of principal components, and only factors corresponding to an eigenvalue >1 were retained for analysis. Varimax rotations were conducted to facilitate the interpretation of identified factors. Variables with factor loading 0.45 (or -0.45) were evaluated for the interpretation of each factor. Data on mortality were collected from baseline to 1996 follow-up; a copy of the death certificate was obtained for each participant died. Factor scores, that are combinations of the original variables and represent the predicted value of a specific factor, were then used as independent variables in Cox proportional hazards models, adjusted for age, to determine the risk of death attributable to each factor. The assumption of proportionality was assessed through the analysis of Schoenfeld residuals of the covariates introduced in the model. Adjusted hazard ratio (HR) and 95% confidence intervals (CIs) were calculated to estimate the strength of this association. All statistical analyses were performed using SAS, version 9.1.3 package (Cary, SAS Institute).

Results Baseline characteristics Table 1 presents the demographic and clinical characteristics of the sample, stratified by sex and study groups. The mean ages of NFG, IFG and T2D subjects were not significantly different, for both sexes, as well as the percentage of actual smokers. The percentage of subjects with hypertension increased from NFG subjects (55.0% among men, 64.4% among women), over IFG subjects (64.1% among men, 72.4% among women), to subjects with T2D (69.8% among men, 77.7% among women; p for trend <0.0001 and 0.001, respectively). No trend was found in relation to percentage of men with heart failure, myocardial infarction, stroke or angina, while among women there was a significant and increasing trend only for heart failure (6.4% for NFG, 6.1% for IFG, 15.5% for T2D subjects, p for trend Z 0.0045) and angina (5.9% for NFG, 6.2% for IFG, 11.9% for T2D subjects, p for trend Z 0.0336). Table 1 shows that, for both men and women, fasting insulin, fasting plasma glucose, HOMA IR and triglycerides,

M. Noale et al. steadily and significantly increased from NFG subjects, over IFG subjects to T2D elderly, while Apo A-I steadily and significantly decreased. LDL levels steadily and significantly decreased among women, while it increased among men from NFG, over IFG to T2D subjects. Only for men, total cholesterol and Apo B significantly increased from NFG, over IFG to T2D participants (p for trend Z 0.0001 and <0.0001, respectively), while HDL-C decreased (48.3  11.9 mg/dl for NFG, 46.1  8.0 mg/dl for IFG, 46.2  8.2 for T2D subjects, p for trend Z 0.0209). A significant increasing trend was found also for BMI and waist circumference, from NFG, over IFG to T2D subjects, for both men and women (p < 0.0001).

Factor analysis Factor analyses were stratified by sex and reduced 10 variables into four factors, both among men and women. Factor loadings after varimax rotation are presented in Table 2. In men, the first factor correlates with HOMA IR, fasting insulin and fasting glucose, and was interpreted as “insulin resistance”; the second factor correlates with BMI and waist circumference (“body size” factor). The third factor correlates with total cholesterol, Apo B and triglycerides, and could represent the “total cholesterol” factor, while the fourth with Apo A-I and HDL-C (“HDL-C” factor). These four factors accounted for a cumulative percentage of 69.2% of the total variance of the original variables. In women, HOMA IR, fasting insulin and fasting glucose clustered as factor 1 (“Insulin resistance”); APO B, total cholesterol and triglycerides, as factor 2 (“Total cholesterol”); BMI and waist circumference as factor 3 (“Body size”); Apo A-I and HDL-C as factor 4 (“HDL-C”). These four factors accounted for a cumulative percentage of 73.8% of the variance in the original variables.

Prospective study During the follow-up (mean time 3.4  0.6 years), deaths were recorded in total population and in subpopulation, and are presented in Table 3. Among men, 149 fatal events were recorded; the overall incidence of fatal events was 15.7% for men with NFG, 11.3% for men with IFG and 16.6% for men with T2D. 139 fatal events were observed among elderly women. The overall incidence of total fatal events for women increased from 8.8% in NFG, 14.1% in IFG and 17.5% in T2D subjects. The “HDL-C” factor was significantly associated with fatal endpoints in men with NFG, IFG and T2D (Table 4a; HR Z 0.79 95% CI 0.67e0.93; HR Z 0.59, 95% CI 0.45e0.79; HR Z 0.55, 95% CI 0.34e0.89, respectively), and in women with T2D (Table 4b; HR Z 0.61, 95% CI 0.44e0.86). No significant interactions with time were found. Among NFG women also the “Body size” factor was a significant protective factor for fatal endpoints (HR Z 0.74, 95% CI 0.57e0.95).

Discussion This study confirms that a “factor including low Apo A-1 and low HDL-C” is a risk factor for all-cause mortality in older men, independently from the glycemia level, and in women

Lipid risk factors among elderly with normal fasting glucose, impaired fasting glucose and type 2 diabetes mellitus

223

Table 1 Baseline characteristics of subjects with NFG, IFG, type 2 diabetes. Data are means  SD, unless otherwise indicated. ILSA, weighted data. Men

Age (years) Smoking status (actual smoker %) Hypertension (%) Heart Failure (%) Myocardial Infarction (%) Stroke (%) Angina (%) Fasting insulin (mU/ml) Fasting plasma glucose (mg/dl) HOMA IR Triglycerides (mg/dl) Total Cholesterol (mg/dl) LDL Cholesterol (mg/dl) HDL Cholesterol (mg/dl) APO A1 (mg/dl) APO B (mg/dl) Body Mass Index (BMI) (kg/m2) Waist circumference (cm)

Women

NFG (n Z 612)

IFG (n Z 341)

T2D (n Z 136)

p-value

NFG (n Z 811)

IFG (n Z 356)

T2D (n Z 166)

p-value

71.9  5.0 20.2

71.0  4.8 20.1

71.2  4.7 20.7

nsb nsa

72.1  6.1 10.6

71.9  6.1 9.3

73.1  6.0 4.6

nsb nsa

55.0 5.7 11.6

64.1 4.8 9.6

69.8 6.1 21.1

<0.0001a 64.4 nsa 6.4 nsa 5.6

72.4 6.1 5.1

77.7 15.5 7.1

0.001a 0.0045a nsa

5.9 7.4 16.5  10.5

5.7 8.3 21.3  15.2

7.0 12.6 25.6  18.2

nsa 5.1 nsa 5.9 <0.0001b 17.0  15.2

4.8 6.2 21.6  14.5

7.6 11.9 22.5  16.4

nsa 0.0336a <0.0001b

89.8  6.5

111.7  9.8

162.9  60.2 <0.0001b 88.7  8.2

111.8  12.2 157.8  53.5 <0.0001b

3.7  2.4 5.9  4.3 10.4  7.9 <0.0001b 3.7  3.2 6.0  4.3 8.7  6.9 <0.0001b b 136.3  52.8 158.9  66.2 172.5  77.7 <0.0001 140.1  62.4 152.0  70.0 173.1  79.3 <0.0001b 205.3  37.1 215.1  36.1 215.2  37.5 0.0001b

230.6  45.8 232.4  49.6 222.4  46.2 nsb

129.7  37.8 132.2  34.0 134.5  36.8 0.0197b

153.5  46.8 151.7  50.4 138.5  46.1 0.0005b

48.3  11.9

49.1  13.7

46.1  8.0

46.2  8.2

0.0209b

50.3  12.9

49.2  10.6

nsb

148.6  24.7 146.2  26.1 141.3  20.1 0.0027b 164.9  31.4 163.9  33.7 156.1  29.9 0.0057b 129.0  59.1 136.1  39.3 144.1  74.2 <0.0001b 134.9  39.5 137.5  43.4 138.1  43.3 nsb 28.7  5.9 28.3  5.1 <0.0001b 25.7  3.2 27.2  3.8 27.1  3.7 <0.0001b 26.9  5.5 95.3  9.5

99.5  10.4

99.8  10.3

<0.0001b 95.2  14.3

99.5  14.8

100.4  12.1 <0.0001b

ns, not significant. a Differences for categorical variables: Mantel-Haenszel chi-square test for trend (not weighted data). b Differences for continuous variables: ordered JonckheereeTerpstra test.

with T2D. We know that HDL-C concentrations decrease in males during puberty and early adulthood, and thereafter remain lower than those in women. This trend could explain why low HDL-C level is a risk factor for mortality in men, independently from other risk factors [21], while in women, at advanced age, is a risk factors only in those with T2D and, therefore, with a cluster of other risk factors. HDL-C in adults was shown to decrease with age in both men and women in prospective studies [22,23]. However, decline in HDL concentration occurs with age because of hormonal changes and inflammatory processes, and it starts later in women than in men [24]. Manolio et al. [25] assessed the relationship between low levels of HDL-C and CHD mortality in 19 455 individuals, and found that middle-aged individuals with low levels of HDL-C versus high levels of HDL-C represented an approximate two-fold increase in the risk of CHD mortality, although the association was not significant in older individuals. To the contrary, the PSC [26] study, in an analysis

of 153 798 individuals, found a 32e37% decrease in the risk of ischemic heart disease (IHD) mortality associated with a 1 SD increase in HDL-C, with similar results found in middle-aged and elderly individuals. Furthermore, PSC tested for interaction effects between HDL-C and other lipids (total cholesterol, triglyceride, and non-HDL-C) and found that the significant associations among HDL-C and CVD fatal and non fatal events were consistent across other lipid risk factors. Our results are in contrast with those reported in a recent review [27], showing a consistent inverse association between HDL-C and CVD risk in both men and women, independently from age, although a weaker association was reported among men than women. The discrepancy with previous studies could be due to the fact that in our study we are considering a “factor including both HDL-C and Apo A-I”. While, as we have seen, there is almost universal consensus on the evidence of a protective effect of HDL-C on CVD, there are still some

224

M. Noale et al.

Table 2

Factor analysis after varimax rotation. ILSA, weighted data. Men

Women

Factor 1 Factor 2 Factor 3 Factor 4 Factor 1 Factor 2 Factor 3 Factor 4 “Insulin “Body Size” “Total “HDL “Insulin “Total “Body Size” “HDL resistance” cholesterol” cholesterol” resistance” cholesterol” cholesterol” Triglycerides 0.24 Total Cholesterol 0.05 HDL Cholesterol 0.00 APO A1 0.12 APO B 0.05 Fasting Insulin 0.88 HOMA IR 0.98 Fasting glucose 0.58 BMI 0.14 Waist 0.15 circumference % of variance 28.8

0.17 0.01 0.03 0.13 0.06 0.14 0.12 0.05 0.94 0.93

0.60 0.81 0.28 0.28 0.68 0.03 0.06 0.12 0.05 0.08

0.35 0.25 0.66 0.77 0.09 0.00 0.04 0.08 0.05 0.06

0.22 0.06 0.08 0.13 0.05 0.87 0.97 0.55 0.13 0.14

0.57 0.88 0.27 0.20 0.90 0.00 0.01 0.06 0.02 0.00

0.16 0.06 0.04 0.05 0.01 0.03 0.10 0.15 0.94 0.94

0.43 0.16 0.67 0.80 0.13 0.17 0.11 0.12 0.07 0.06

15.4

13.8

11.2

21.2

20.3

18.3

14.0

In bold, factor loadings 0.45 or 0.45.

uncertainties about the mechanism in which HDL-C exerts its effect. Apo A-I was not included in most of the previous studies. It is well known that the lipoprotein HDL has two important roles: first, it promotes reverse cholesterol transport, and second, it modulates inflammation. Epidemiological studies show that HDL-C levels are inversely correlated with the risk of cardiovascular events. However, many patients who experience a clinical event have normal, or even high, levels of HDL-C. Measuring HDL-C levels provides information about the size of the HDL pool, but does not predict HDL composition or function [28]. The main component of HDL, apolipoprotein A-I (Apo A-I), is largely responsible for reverse cholesterol transport through the macrophage. Recent reviews have reported that patients with CHD had substantially lower Apo A-I concentrations than do healthy controls. More recently, the

Table 3 Deaths from baseline to 1996 follow-up, in the total population and in subpopulations. ILSA, weighted data.

Men Women

Table 4a

Factor Factor Factor Factor

1 2 3 4

Total population

NFG

IFG

T2D

149 (14.4%) 139 (11.3%)

91 (15.7%) 66 (8.8%)

37 (11.3%) 46 (14.1%)

20 (16.6%) 26 (17.5%)

INTERHEART case-control study, which involved approximately 15,000 myocardial infarction cases and 15,000 controls from 52 countries, reported that Apo B/Apo A-I was one of the strongest correlates of prevalent disease amongst the several anthropometric, psychosocial and established risk factors measured. In our sample, we have shown a significant decline of Apo A-I from normoglycemic individuals to T2D, in both sexes, and this could support a stronger, negative effect of low Apo A-I compared to HDL-C. The factor related to “body size” in women was almost completed saturated by BMI and waist circumference. This factor was significantly protective from all-cause mortality only among NFG women. It has been reported that higher BMI values indicated a lower mortality risk once the risk attributable to waist circumference was accounted for [29]. Lean mass in overweight/obese older individuals may act as a nutritional reserve against traumatic events and protect from mortality [30,31]. Bigaard et al. [32] suggested that BMI can be considered as a reflection of lean mass in individuals with the same waist circumference, and waist circumference as a reflection of total-abdominal fat content in individuals with the same BMI; this observation suggests that, after consideration of waist circumference, greater BMI may represent a unique aspect of body composition e one that decreases health risk [29].

Cox models for fatal events among men, in the total population and in subpopulations: HR and (95% CI). ILSA.

“Insulin resistance” “Body size” “Total cholesterol” “HDL cholesterol”

Total population

NFG

1.07 0.93 0.91 0.72

0.86 0.91 0.84 0.79

Models are controlled also for age. In bold, significant protective factors.

(0.95e1.21) (0.82e1.06) (0.79e1.04) (0.63e0.82)

IFG (0.62e1.18) (0.77e1.08) (0.70e1.02) (0.67e0.93)

1.18 0.95 1.11 0.59

T2d (0.96e1.46) (0.74e1.22) (0.87e1.43) (0.45e0.79)

1.17 0.87 0.66 0.55

(0.97e1.42) (0.57e1.32) (0.43e1.01) (0.34e0.89)

Lipid risk factors among elderly with normal fasting glucose, impaired fasting glucose and type 2 diabetes mellitus Table 4b

Factor Factor Factor Factor

1 2 3 4

225

Cox models for fatal events among women, in the total population and in subpopulations: HR and (95% CI). ILSA.

“Insulin resistance” “Total cholesterol” “Body size” “HDL cholesterol”

Total population

NFG

1.09 0.94 0.81 0.89

0.78 0.83 0.74 0.89

(0.94e1.27) (0.82e1.07) (0.68e0.95) (0.78e1.03)

IFG (0.53e1.16) (0.69e1.01) (0.57e0.95) (0.73e1.09)

1.22 1.02 0.74 1.15

T2D (0.84e1.79) (0.82e1.27) (0.55e1.01) (0.90e1.48)

1.08 0.84 0.97 0.61

(0.72e1.62) (0.56e1.24) (0.66e1.42) (0.44e0.86)

Models are controlled also for age. In bold, significant protective factors.

Our study has some limitations. First, we didn’t use the latest criteria for the diagnosis of diabetes [20], because the prevalence rates of the diseases were calculated in 1994 [16]. Therefore, we have classified as hyperglycemic individuals those not previously diagnosed and with fasting glycemia between 126 mg/dl and 139 mg/dl. Given this potential misclassification, we might have reported a stronger association between low HDL-C and Apo A-I and total mortality in hyperglycemic individuals. However, even excluding individuals with glycemia 126 mg/dl, who didn’t undergo the second phase for the clinical confirmation of diabetes, we found results similar to those reported here. Second, we stratified analyses by sex, adjusted for age and for other factors identified by factor analysis, but residual confounding by imperfectly measured or unmeasured confounders cannot be excluded. However, this is a common limitation of observational studies. Third, despite the dimension of the ILSA sample size at the baseline, the number of deaths at follow-up in each group is small. Studies on larger samples with longer follow-up are needed to confirm the results. The strengths of this study include its population-based design with a long follow-up, a large sample representative of the Italian older population, the clinical diagnoses of diseases and the reliable assessment of metabolic and cardiovascular risk factors and of causes of death. Future studies are required to find strategies to mimic the positive effects of HDL-C and of Apo A-I and to prevent the negative effects of aging on these cardioprotective factors in the aged population. Our study demonstrate that low HDL-C and Apo A-I levels represent an independent risk factor for overall mortality in both men and women with diabetes and also in men with normo- and hyperglycemia. Current therapeutic use of statins as monotherapy is still leaving many patients with mixed atherogenic dyslipidemia at high risk for coronary events and different, more comprehensive approaches, including life style changes, could be appropriate in these patients.

Conflict of interest We certify that there is no conflict of interest regarding the material discussed in the manuscript.

Acknowledgments The ILSA Working Group

E. Scafato, MD (scientific coordinator), G. Farchi, MSc, L. Galluzzo, MA, C. Gandin, MD, Istituto Superiore di Sanita `, Roma; A. Capurso, MD, F. Panza, MD, PhD, V. Solfrizzi, MD, PhD, V. Lepore, MD, P. Livrea, MD, University of Bari; L. Motta, MD, G. Carnazzo, MD, M. Motta, MD, P. Bentivegna, MD, University of Catania; S. Bonaiuto, MD, G. Cruciani, MD, D. Postacchini, MD, Italian National Research Center on Aging, Fermo; D. Inzitari, MD, L. Amaducci, MD, University of Firenze; A. Di Carlo, MD, M. Baldereschi, MD, CNR, Firenze; C. Gandolfo, MD, M. Conti, MD, University of Genova; N. Canal, MD, M. Franceschi, MD, San Raffaele Institute, Milano; G. Scarlato, MD, L. Candelise, MD, E. Scapini, MD, University of Milano; F. Rengo, MD, P. Abete, MD, F. Cacciatore, MD, University of Napoli; G. Enzi, MD, L. Battistin, MD, G. Sergi, MD, G. Crepaldi, MD, University of Padova; S. Maggi, MD, N. Minicuci, PhD, M. Noale, ScD, CNR, Aging Section, Padova; F. Grigoletto, ScD, E. Perissinotto, ScD, Institute of Hygiene, University of Padova; P. Carbonin, MD, Universita ` Cattolica del Sacro Cuore, Roma. The Italian National Research Council supported the ILSA project from 1991 to 1998 as part of the Progetto Finalizzato Invecchiamento. Since 1999, the Italian National Research Council, the “Biology of Aging” Strategic Project and the Ministero della Sanita `, through the program “Epidemiology of the Elderly” of the Istituto Superiore di Sanita ` and the “Estimates of Health Needs of the Elderly” Special Programme of the Tuscany Region have been supporting the ILSA.

References [1] Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes, other risk factors, and 12-yr cardiovascular mortality for men screened in the multiple risk factor intervention trial. Diabetes Care 1993;6:434e44. [2] Lowe LP, Liu K, Greenland P, Metzger BE, Dyer AR, Stamler J. Diabetes, asymptomatic hyperglycaemia and 22-year mortality in black and white men: the Chicago Heart Association detection project in industry study. Diabetes Care 1997;20:163e9. [3] Wei M, Gaskill SP, Haffner SM, Stern MP. Effects of diabetes and level of glycaemia on all-cause and cardiovascular mortality: the San Antonio Heart Study. Diabetes Care 1998; 21:1167e72. [4] Gu K, Cowie CC, Harris MI. Mortality in adults with and without diabetes in a national cohort of the US population, 1971e1993. Diabetes Care 1998;21:1138e45. [5] Morgan CL, Currie CJ, Peters JR. Relationship between diabetes and mortality: a population study using record linkage. Diabetes Care 2000;23:1103e7. [6] Barr ELM, Zimmet PZ, Welborn TA, Jolley D, Magliano DJ, Dunstan DW, et al. Risk of cardiovascular and all-cause

226

[7]

[8]

[9]

[10]

[11]

[12]

[13] [14]

[15]

[16]

[17]

[18]

M. Noale et al. mortality in individuals with diabetes mellitus, impaired fasting glucose and impaired glucose tolerance. The Australian diabetes, obesity and lifestyle study (AusDiab). Circulation 2007;116:151e7. Coutinho M, Gerstein HC, Wang Y, Yusuf S. The relationship between glucose and incident cardiovascular events: a metaregression analysis of published data from 20 studies of 95,783 individuals followed for 12.4 years. Diabetes Care 1999;22: 233e40. The DECODE Study Group on behalf of the European Diabetes Epidemiology Group. Glucose tolerance and cardiovascular mortality: comparison of fasting and 2-hour diagnostic criteria. Arch Intern Med 2001;161:397e405. Nakagami T. Hyperglycaemia and mortality from all causes and from cardiovascular disease in five populations of Asian origin. Diabetologia 2004;47:385e94. Florkowski CM, Scott RS, Coope PA, Moir CL. Predictors of mortality from T2D in Canterbury, New Zealand; a ten-year cohort study. Diabetes Res Clin Pract 2001;21:113e20. Florkowski CM, Scott RS, Moir CL, Graham PJ. Lipid but not glycaemic parameters predict total mortality from type 2 diabetes mellitus in Canterbury, New Zealand. Diabetic Med 1998;15:386e92. Drexel H, Aczel S, Marte T, Benzer W, Langer P, Moll W, et al. Is atherosclerosis in diabetes and impaired fasting glucose driven by elevated LDL cholesterol or by decreased HDL cholesterol? Diabetes Care 2005;28:108e14. Betteridge DJ. Diabetic dyslipidaemia. Eur J Clin Invest 1999; 29:12e6. Howard BV, Knowler WC, Vasquez B, Kennedy AL, Pettitt DJ, Bennett PH, et al. Plasma and lipoprotein cholesterol and triglyceride in the Pima Indian population: comparison of diabetics and nondiabetics. Arteriosclerosis 1984;4:462e71. Collins R, Armitage J, Parish S, Sleigh P, Peto R. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection study of cholesterol-lowering with simvastatin in 5963 people with diabetes: a randomised placebo-controlled trial. Lancet 2003;361:2005e16. Maggi S, Zucchetto M, Grigoletto F, Baldereschi M, Candelise L, Scarpini E, et al. for the ILSA group. The Italian longitudinal study on aging (ILSA): design and methods. Aging Clin Exp Res 1994;6:464e73. The Italian Longitudinal Study on Aging Working Group. Prevalence of chronic diseases in older Italians: comparing selfreported and clinical diagnoses. Int J Epidemiol 1997;26: 995e1002. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low density lipoprotein cholesterol in plasma

[19]

[20]

[21] [22]

[23]

[24] [25]

[26]

[27]

[28]

[29]

[30]

[31] [32]

without use of the preparative ultracentrifuge. Clin Chem 1972;18:499e502. Perissinotto E, Pisent C, Sergi G, Grigoletto F, Enzi G. for the ILSA Working Group. Anthropometric measurements in the elderly: age gender differences. Br J Nutr 2002;87:1e11. The Expert Committee on the diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the diagnosis and Classification of diabetes mellitus. Diabetes Care 1997;20:1183e97. Kreisberg RA, Kasim S. Cholesterol metabolism and aging. Am J Med 1987;82:54e60. Wilson PWF, Anderson KM, Harris T, Kannel WB, Castelli WP. Determinants of change in total cholesterol and HDL-Ch with age: the Framingham study. J Gerontol 1994;49:M252e7. Ferrara A, Barrett-Connor E, Shan J. Total, LDL-Ch, and HDLCh decrease with age in older men and women the Rancho Bernardo study 1984e1994. Circulation 1997;96:37e43. Walter M. Interrelationships among HDL metabolism, aging, and atherosclerosis. Arterioscler Thromb Vasc Biol; 2009:1244e50. Manolio TA, Pearson TA, Wenger NK, Barrett-Connor E, Payne GH, Harlan WR. Cholesterol and heart disease in older persons and women. Review of an NHLBI workshop. Ann Epidemiol 1992;2(1e2):161e76. Lewington S, Whitlock G, Clarke R, Sherliker P, Emberson J, Halsey J, et alProspective Studies Collaboration. Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths. Lancet 2007;370(9602):1829e39. Chirovsky DR, Fedirko V, Cui Y, Sazonov V, Barter P. Prospective studies on the relationship between high-density lipoprotein cholesterol and cardiovascular risk: a systematic review. Eur J Cardiovasc Prev Rehabil 2009;16(4):404e23. Navab M, Reddy ST, Van Lenten BJ, Fogelman AM. HDL and cardiovascular disease: atherogenic and atheroprotective mechanisms. Nat Rev Cardiol; 2011 Feb 8 [Epub ahead of print]. Janssen I, Katzmarzyk PT, Ross R. Body mass index is inversely related to mortality in older people after adjustment for waist circumference. J Am Geriatr Soc 2005 Dec;53(12):2112e8. Heiat A, Vaccarino V, Krumholz HM. An evidence-based assessment of federal guidelines for overweight and obesity as they apply to elderly persons. Arch Intern Med 2001 May 14; 161(9):1194e203. Elia M. Obesity in the elderly. Obes Res 2001 Nov;9(Suppl. 4): 244Se8S [Review]. Bigaard J, Tjønneland A, Thomsen BL, Overvad K, Heitmann BL, Sørensen TI. Waist circumference, BMI, smoking, and mortality in middle-aged men and women. Obes Res 2003 Jul;11(7):895e903.