Eur J Nutr (2013) 52:307316

DOI 10.1007/s00394-012-0337-1

ORIGINAL CONTRIBUTION

Metabolic syndrome and quality of life in the elderly:

age and gender differences
Alice Laudisio Emanuele Marzetti Livia Antonica Francesco Pagano
Davide L. Vetrano Roberto Bernabei Giuseppe Zuccala`

Received: 7 September 2011 / Accepted: 24 February 2012 / Published online: 11 March 2012
Springer-Verlag 2012

Abstract
Purpose The metabolic syndrome (MetS) is highly prevalent in Western older populations. MetS is an intriguing
entity, because it includes potentially reversible risk factors. Some studies have suggested an inverse correlation
between MetS and health-related quality of life (HRQoL),
but data regarding older subjects are scanty and conflicting.
The aim of this study was to assess the association between
HRQoL and MetS in older, unselected community-dwelling subjects.
Methods We analyzed data of 356 subjects aged 75?
living in Tuscania (Italy). HRQoL was assessed using the
Health Utilities Index, Mark 3. Diagnosis of MetS was
defined according to the National Cholesterol Education
Programs ATP-III criteria.
Results MetS was reported by 137 (38%) participants.
According to linear regression analysis, MetS was associated with significantly better HRQoL in men (B = 0.19
95% CI = 0.060.32; p = 0.006), but not in women. Also,
when the regression model was analyzed in men, MetS
was associated with better HRQoL (B = 0.17, 95%
CI = 0.010.32; p = 0.035) only among participants aged
80?. No significant associations were found in men
between HRQoL and any of the single components of
MetS.
Conclusions MetS is not associated with worse HRQoL
among community-dwelling elderly; it is associated with
significantly better HRQoL among the oldest men.

A. Laudisio (&)
E. Marzetti
L. Antonica
F. Pagano

D. L. Vetrano
R. Bernabei
G. Zuccala`
Department of Gerontology and Geriatrics, Catholic University
of Medicine, L.go F. Vito, 1, 00168 Rome, Italy
e-mail: [email protected]

Keywords Metabolic syndrome
Quality of life

Elderly
Epidemiology

Introduction
The metabolic syndrome (MetS) is defined by abdominal
adiposity, elevated triglyceride level, low high-density
lipoprotein cholesterol level, high blood pressure, and high
fasting blood glucose [1]. This syndrome is increasingly
being recognized in geriatric populations [2]; the pathophysiological pathway seems to involve hyperinsulinemia,
which is strongly associated with obesity. The mechanisms
that link obesity with hyperinsulinemia, in turn, are thought
to involve primarily insulin resistance, which is chiefly
attributed to adipose tissue inflammation and dysfunction
with ensuing release of inflammatory cytokines [3]. Not
surprisingly, MetS has been associated with increased risk
of diabetes and accelerated atherosclerosis, as well as with
its complications, including stroke and myocardial infarction [4, 5]. Thus, recognition and treatment of the abovementioned components of MetS are recommended for
prevention at the individual, as well as population levels
[6].
On the other hand, it is acknowledged that health-related
quality of life (HRQoL) as perceived by the patient should
be considered the main outcome of any health interventions. Studies so far conducted generally found an association of MetS, as well as its components, with worse
HRQoL [7]; in some studies, however, MetS has been
found to be associated with reduced HRQoL in women, but
not in men [8, 9]. However, these studies generally enrolled
selected, middle-aged populations, often with no mention
of comorbidity. Little is known about elderly populations,
especially the oldest [10]. This issue is relevant for the

123

308

implementation of preventive services at the population

level, as oldest age subjects, among whom the components
of MetS are most prevalent, are steeply increasing in
Western countries. Also, it has been found that the costeffectiveness of preventive interventions for type II diabetes mellitus is strongly dependent upon the HRQoL
perceived by subjects with MetS [6].
The aim of the present study was to evaluate the association of MetS with self-assessed quality of life in an
unselected, community-dwelling older population, and to
ascertain whether such an association varies according to
sex and age.

Methods
Participants
This study involved 356 subjects aged C75 years who were
living in Tuscania (Italy) on January 1st, 2004. All 387
inhabitants of this town aged 75? had been enrolled in a
study of the genetic determinants of health status in six
Italian towns and were visited by the study researchers
between January 1st and June 1st, 2004. For the present
study, we excluded 31 of 387 subjects because of missing
data for the study variables. All participants underwent
ambulatory or home visits by the study physicians, who
performed physical examination, electrocardiography,
Doppler echocardiography, ultrasonographic bone densitometry, and collected medical history and blood samples
for serum chemistry and genomic analyses. Also, they
completed a questionnaire that included participants data
on socioeconomic status and lifestyle habits. Appointments
dedicated to blood sample collection were given early in
the morning.
The Institutional Review Board approved the protocol
of the present study, and all patients provided written
informed consent.
Covariates
Education was expressed as years of school attendance.
Alcohol consumption was defined by consuming at least
two drinks per week. Drinks were recorded as wine
units (100 mL), because this beverage represents by
far the major form of alcohol consumption in Italy,
independently of any seasonal variations [11]. A conversion table was used for other alcoholic beverages;
each liter of wine was assumed to contain 80 g of
alcohol [12].
Smoking was calculated as total lifetime pack years for
current, as well as former smokers. Income was expressed
as perceived adequate or inadequate by the participant.

123

Eur J Nutr (2013) 52:307316

Functional ability was estimated using the Katz activities of daily living (ADLs) [13], and the Lawton and
Brody scale for instrumental activities of daily living
(IADLs) [14]. We adopted the same scoring for women and
men (i.e., 08) for the IADLs because men had a median
number of preserved instrumental activities of 7. This
figure is higher than the maximum score of 5 that has been
sometimes adopted for men; furthermore, it was not different from that reported by older female participants
(median = 7). Depressive symptoms were evaluated using
the validated 30-item Italian version of the Geriatric
Depression Scale (GDS) [15], a self-reported scale based
upon yes-or-no questions regarding mood over the previous
week. A cutoff of 11 is generally adopted to diagnose
depression. The test yields 84% sensitivity and 95%
specificity for the diagnosis of depression [16]. Cognitive
performance was assessed using the Hodkinson abbreviated mental test (AMT) [17], which has been validated for
detection of cognitive impairment in older, including Italian, populations [18]. A major advantage of this scale is
represented by the lack of written items, which are known
to bias the assessment of cognitive function in older populations with prevalent illiteracy, such as that of the present
study.
Participants diagnoses and drug treatments were
obtained by their general practitioners and further confirmed by the study physicians, who received specific
training and whose concordance had been tested (Cohens
Kappa [0.80 for all the five proposed dummy cases).
All drugs assumed by participants were coded according
to the Anatomical Therapeutic and Chemical codes.
Diagnoses were coded according to the International
Classification of Diseases, ninth edition, Clinical Modification codes [19]. Comorbidity was quantified using the
Charlson comorbidity index score by adding scores
assigned to specific diagnoses [20].
Blood samples were obtained after overnight fast; the
processed specimens were aliquoted into cryovials, frozen
at -70 C, and shipped to the Department of Experimental
Pathology, University of Bologna. Serum creatinine was
measured by a standard creatinine Jaffe method (Roche
Diagnostics, Mannheim, Germany). Hemoglobin was
measured using the hematology automated Autoanalyzer
DASIT SE 9000 (Sysmex Corporation, Kobe, Japan).
Albumin was measured using an agarose electrophoretic
technique (Hydragel Protein(E) 15/30; Sebia, Issy-lesMoulineaux, France). Sodium was measured by indirect
ion-selective electrode (ISE) method (DXC600, Beckman
Coulter, Brea, USA).
Measurements of high-sensitivity C-reactive protein
were performed in duplicate by enzyme-linked immunosorbent assay (ELISA). These objective tests were analyzed as potential confounders of the association between

Eur J Nutr (2013) 52:307316

MetS and HRQoL, because they reflected relevant conditions that might be affected by MetS and might influence
HRQoL, namely renal function (serum creatinine), nutritional status (albumin, hemoglobin), hydration status
(sodium), and subclinical inflammation (high-sensitivity
C-reactive protein).
Perceived quality of life
Perceived health-related quality of life was assessed using
the Health Utilities Index, Mark 3 (HUI3) [21]. This tool
calculates self-assessed, preference-based quality of life
using a 15-item questionnaire. The HUI3 explores eight
domains, namely vision, hearing, speech, ambulation,
dexterity, emotion, cognition, and pain. A overall utility
score is calculated as the value of a single function that
composition is based upon preference weights for 243
health states obtained from representative populations from
several countries. HUI3 scores range from -0.36 to 1.00.
Higher scores indicate better quality of life; the value of 1
corresponds to the best conceivable health status and zero
to instant death, negative scores representing states considered worse than death. The utility scores calculated
by this multi-attribute health-status classification system
allow to measure the overall HRQL for patients, which is
also used in formal cost-utility and cost-effectiveness
analyses.
Metabolic syndrome
The metabolic syndrome was defined according to the
National Cholesterol Education Programs ATP-III criteria,
adding use of hypolipemic, hypoglycemic, and antihypertensive medications, as already done in several previous
studies; the diagnosis of the metabolic syndrome was
established as the presence of three or more of the following features: waist circumference [88 cm in women
and [102 cm in men; fasting serum triglycerides C150
mg/dL; serum HDL \50 mg/dL in women and \40 mg/dL
in men; blood pressure C130/85 mmHg; fasting blood
glucose levels C110 mg/dL [1]. Waist circumference
was measured at the midpoint between the lower rib
margin and the iliac crest (normally umbilical level).
Blood pressure was recorded using a standard mercury
sphygmomanometer.
All blood pressure measurements were performed in
supine position by three measurements, separated by 2-min
intervals; the average of the last two measures was used in
the analyses. HDL cholesterol and triglycerides concentrations were determined using commercial enzymatic tests
(Roche Diagnostics, Mannheim, Germany). Serum glucose
levels were determined by enzymatic colorimetric assay
(Roche Diagnostics, Mannheim, Germany).

309

Statistical analyses
Data of continuous variables are presented as mean values standard deviation (SD). Statistical analyses were
performed using Statistical Package for the Social Sciences
(SPSS for Windows version 13.0, 2004, SPSS Inc.,
Chicago, IL); differences were considered significant at
the p \ 0.05 level. Analysis of variance (ANOVA) for
normally distributed variables according to MetS was
performed by ANOVA comparisons; otherwise, the nonparametric KruskalWallis H test was adopted. Chi-square
analysis was used for dichotomous variables. Serum
C-reactive protein levels were analyzed after log transformation; smoking and the Health Utilities Index score were
not transformed because the tests for linearity and deviation from linearity indicated that the association of these
variables with MetS had significant F ratio for linearity
(p \ 0.02 for both variables), but not for nonlinear components (p [ 0.5 for both variables).
Linear regression analysis was used to estimate the
association of variables of interest, including MetS, with
the HUI3 score. To assess independent correlates of
HRQoL, which might confound the association between
the HUI3 score and MetS, groups of variables (demographics, comorbid conditions, medications, and objective
tests, as depicted in Table 1) were first examined using
separate age- and sex-adjusted regression models (Table 2,
left columns). Those variables, significant at the p \ 0.050
level in these initial models, were simultaneously entered
into a summary model (Table 2, right columns). The
summary model was anyway adjusted for age and sex, as
generally recommended for more conservative analyses.
The same fully adjusted regression model was analyzed
after stratifying for sex (Table 3). Also, the fully adjusted
regression model was analyzed in logistic regression
analysis considering a HUI3 score above or below the
median value, among male participants (Table 4).
Analyses of the interaction term MetS*consumption of
alcohol, MetS*use of angiotensin-converting enzyme
inhibitors, and MetS*use of benzodiazepines were also
performed to assess whether the association of MetS with a
HUI3 score above the median value varied according to
consumption of alcohol, use of angiotensin-converting
enzyme inhibitors, or benzodiazepines, also after stratifying for sex.
In addition, the association between the HUI3 score and
MetS was assessed in men after stratifying for age \ or
C80 years (i.e., the median age of the study sample), using
the fully adjusted linear regression model (Table 5).
Analysis of the interaction term MetS*age was performed using the fully adjusted logistic regression model to
ascertain whether the association between MetS and HUI3
varied according to age. Also, the fully adjusted linear

123

310

Eur J Nutr (2013) 52:307316

Table 1 Characteristics of participants according to diagnosis of the metabolic syndrome

Participants with
metabolic syndrome (n = 137)
n (%) or mean SD

Participants without
metabolic syndrome (n = 219)
n (%) or mean SD

Age (years)

79 5

79 6

0.973a

Sex (female)
Education (years)

88 (64%)
43

106 (48%)
53

0.004b
0.101a

Current alcohol consumption

94 (69%)

155 (72%)

0.004b

Smokinga

3,909 7,781

5,437 10,845

0.128c

Incomeb

101 (75%)

171 (81%)

0.221b

Chronic pulmonary disease

39 (28%)

53 (24%)

0.456b

Heart failure

25 (18%)

41 (19%)

0.890b

Arthritis

113 (82%)

164 (76%)

0.184b

Demographics and lifestyle habits

Comorbid conditions

Stroke

15 (11%)

31 (14%)

0.419b

Renal disease

9 (7%)

7 (3%)

0.189b

Depressionc

71(53%)

92 (44%)

0.123b

Hepatic disease

7 (5%)

7 (3%)

0.410b

Cancer

16 (12%)

18 (8%)

0.355b

Charlson comorbidity score index

22

11

0.017a

Medications
SSRId

6 (4%)

7 (3%)

0.573b

Beta-blockers

10 (7%)

7 (3%)

0.123b

Corticosteroids

6 (4%)

8 (4%)

0.783b

ACE-inhibitorse

55 (40%)

51 (23%)

0.001b

Loop diuretics

33 (24%)

36 (16%)

0.098b

Benzodiazepines

37 (27%)

36 (16%)

0.021b

NSAIDSf

8 (6%)

17 (8%)

0.531b

Aspirin

35 (25%)

38 (17%)

0.079 b

Serum creatinine (mg/dL)

1.1 0.3

1.0 0.3

0.056a

Serum albumin (g/dL)

4.2 0.7

4.1 0.5

0.216a

Hemoglobin (g/dL)

14.1 1.6

14.2 1.7

0.913a

Serum sodium (mEq/L)

143 3

144 3

0.325a

C-reactive protein (mg/dL)

0.98 2.02

0.63 1.27

0.087c

Objective tests

Activities of daily living

51

51

0.723a

IADLsg
Hodkinson abbreviated mental test

62
82

62
72

0.930a
0.146a

Quality of lifeh

0.31 0.45

0.30 0.47

0.405a

Total lifetime pack years

Perceived as adequate

Defined by a Geriatric Depression score [11

Selective serotonin reuptake inhibitors

Angiotensin-converting enzyme inhibitors

Nonsteroidal antiinflammatory agents

Instrumental activities of daily living

Measured using the Health Utilities IndexMark 3

a ANOVA comparison
b Chi-square analysis
c Nonparametric KruskalWallis H test

123

Eur J Nutr (2013) 52:307316

311

Table 2 Association (B coefficients, and 95% confidence intervals, CI) of perceived quality of life with the variables of interest according to the
initial (age- and sex-adjusted), and the summary linear (fully adjusted) regression models
Age- and sex-adjusted models
B

95% CI

Summary model
p

95% CI

Demographics and lifestyle habits

0.03

0.02 to 0.04

0.03 to 0.05

\0.0001

Sex (female)
Education (years)

-0.07
0.02

-0.17 to 0.04
0.01 to 0.03

0.235
0.014

-0.12
0.01

-0.21 to -0.03
-0.01 to 0.02

0.006
0.433

Current alcohol consumption

-0.07

-0.13 to -0.01

0.025

-0.029

-0.08 to 0.02

0.285

0.01

-0.03 to 0.08

0.863

-0.08

-0.20 to 0.03

0.151
-0.06

-0.16 to 0.03

0.161

-0.14

-0.23 to -0.05

0.002

Age (years)

Smoking

Incomeb

\0.0001

0.04

Comorbid conditions
Chronic pulmonary disease

-0.12

-0.24 to -0.01

0.043

0.05

-0.09 to 0.18

0.513

-0.01

-0.12 to 0.10

0.900
0.320

Heart failure
Arthritis
Stroke

0.08

-0.08 to 0.24

Renal disease

-0.11

-0.35 to 0.14

Depressionc

-0.27

-0.35 to -0.17

Hepatic disease

-0.11

-0.35 to 0.12

0.349

Cancer

-0.18

-0.37 to 0.01

0.053

Charlson comorbidity score index

0.382
\0.0001

-0.01

-0.06 to 0.04

0.638

Medications
SSRId

0.06

-0.19 to 0.31

0.642

Beta-blockers

0.02

-0.19 to 0.23

0.842

-0.55 to -0.08

0.009

-0.19

-0.41 to 0.03

0.085

0.12

-0.02 to 0.26

0.082

Corticosteroids
ACE-I

-0.31

Loop diuretics

0.24

0.07 to 0.40

0.005

-0.03

-0.15 to 0.09

0.638

Benzodiazepines

0.02

-1.01 to 0.14

0.731

NSAIDSf

-0.02

-0.20 to 0.17

0.846

Aspirin

-0.05

-0.17 to 0.06

0.389

Objective tests
Serum creatinine (mg/dL)

0.01

-0.11 to 0.14

0.824

-0.06

-0.17 to 0.04

0.243

Hemoglobin (g/dL)

0.01

-0.02 to 0.05

0.510

Serum sodium (mEq/L)

0.01

-0.01 to 0.02

0.624

C-reactive protein (mg/dL)g

0.01

-0.05 to 0.05

0.943

ADLh

0.05

-0.00 to 0.11

0.065

IADLi
Hodkinson abbreviated mental test

0.05
0.02

0.02 to 0.08
-0.02 to 0.05

0.003
0.298

0.08

0.06 to 0.10

Metabolic syndrome

0.13

0.02 to 0.25

0.026

0.07

-0.01 to 0.15

Serum albumin (g/dL)

\0.0001
0.072

All the covariates were entered simultaneously into the regression models
a

Total lifetime pack years

Perceived as adequate

30-item Geriatric Depression score [11

Selective serotonin reuptake inhibitors

Angiotensin-converting enzyme inhibitors

Nonsteroidal antiinflammatory agents

High-sensitivity, log-transformed

Activities of daily living

Instrumental activities of daily living

123

312

Eur J Nutr (2013) 52:307316

Table 3 Association (B coefficients, and 95% confidence intervals, CI) between perceived quality of life and the metabolic syndrome according
to the fully adjusted linear regression model, in women and men
Women
B

Men
95% CI

\0.0001

95% CI

p
\0.0001

Age (each year)

0.04

0.03 to 0.05

0.04

0.03 to 0.05

Education (each year)

0.01

-0.01 to 0.03

0.102

-0.01

-0.01 to 0.02

0.943

Current alcohol consumption

Chronic pulmonary disease

-0.04
-0.14

-0.10 to 0.02
-0.27 to -0.02

0.194
0.025

-0.01
0.01

-0.13 to 0.10
-0.14 to 0.14

0.828
0.995

Diagnosis of depressiona

-0.15

-0.27 to -0.04

0.007

-0.09

-0.24 to 0.05

0.185

Corticosteroids

-0.07

-0.33 to 0.19

0.612

-0.27

-0.68 to 0.12

0.176

0.10

-0.06 to 0.25

0.219

0.12

-0.14 to 0.39

0.368

\0.0001

0.08

0.05 to 0.11

\0.0001

0.811

0.19

0.06 to 0.32

0.006

ACE-Ib
IADLc
Metabolic syndrome

0.08

0.06 to 0.11

-0.01

-0.11 to 0.09

All the covariates were entered simultaneously into the regression models
a

30-item Geriatric Depression score [11

Angiotensin-converting enzyme inhibitors

Instrumental activities of daily living

Table 4 Association (odds ratios, OR, and 95% confidence intervals,

CI) between perceived quality of life above the median value (C0.31)
and the metabolic syndrome according to logistic regression in men
OR

95% CI

Age (each year)

1.19

1.091.30

\0.0001

Education (each year)

1.00

0.871.15

0.976

Current alcohol consumption

1.14

0.502.58

0.760

Chronic pulmonary disease

1.00

0.402.50

0.993

Diagnosis of depressiona

0.38

0.151.00

0.040

Corticosteroids

0.43

0.035.65

0.520

ACE-Ib

0.39

0.141.13

0.083

IADLc

1.57

1.281.92

\0.0001

Metabolic syndrome

3.34

1.288.70

0.013

All the covariates were entered simultaneously into the regression

model
a

30-item Geriatric Depression score [11

Angiotensin-converting enzyme inhibitors

Instrumental activities of daily living

regression model was analyzed among men, also after

stratifying for age \ or C80 years, by entering the single
components of MetS (Table 6). Eventually, the variance
inflation factors and condition index were calculated
among men, to assess collinearity between the single
components of MetS (Table 7).

Results
The main characteristics of participants according to the
presence of MetS are depicted in Table 1. Excluded

123

participants did not differ significantly from those included

in the study by age, sex, or prevalent MetS.
MetS was found in 137/356 (38%) participants. Overall,
participants with MetS, as compared with remaining subjects, showed no differences in the HUI3 score (Table 1).
However, men aged 80? with MetS, as compared with
age-matched participants without MetS, had better quality
of life scores (0.57 0.34 vs. 0.39 0.44; p = 0.018).
MetS was more common in women (88/194) than in men
(49/162; Fishers exact test p = 0.004). C-reactive protein
levels did not differ among men and women (0.70 1.36
vs. 0.83 1.82; p = 0.523).
Main characteristics of participants
with and without MetS
Participants with MetS, as compared with other participants, reported less frequent consumption of alcohol
(p = 0.004), and a more prevalent use of angiotensinconverting enzyme inhibitors (p = 0.001) and benzodiazepines (p = 0.021). Noticeably, participants with MetS
had higher Charlson comorbidity score, as compared with
remaining subjects (p = 0.017).
Multivariable analyses
Results of the initial age- and sex-adjusted regression
models for demographics and lifestyle habits, comorbid
conditions, medications, and objective tests, groups of
variables are depicted in the left (age- and sex-adjusted)
columns of Table 2.
In the initial linear regression models, age, sex, MetS,
education level, current alcohol consumption, diagnosis of

Eur J Nutr (2013) 52:307316

313

Table 5 Association (B coefficients, and 95% confidence intervals, CI) between the perceived quality of life score and the metabolic syndrome
according to the fully adjusted linear regression model in men, after stratifying for age
Age \80 years (n = 85)
B

Age C80 years (n = 69)

95% CI

95% CI

Age (each year)

0.09

0.07 to 0.11

\0.0001

-0.01

-0.04 to 0.01

0.231

Education (each year)

0.01

-0.02 to 0.03

0.673

-0.01

-0.02 to 0.02

0.941

Current alcohol consumption

Chronic pulmonary disease

-0.01
0.04

-0.15 to 0.14
-0.15 to 0.22

0.945
0.698

-0.06
0.03

-0.20 to 0.09
-0.13 to 0.19

0.444
0.710

Diagnosis of depressiona

-0.20

-0.39 to -0.01

0.035

-0.15

-0.31 to 0.01

0.064

Corticosteroids

-0.20

-0.65 to 0.26

0.393

0.39

-0.21 to 0.99

0.203

ACE-Ib

0.15

-0.20 to 0.49

0.397

0.09

-0.23 to 0.42

0.559

IADLc

0.06

0.02 to 0.09

0.005

0.06

0.03 to 0.09

0.000

Metabolic syndrome

0.16

-0.02 to 0.35

0.081

0.17

0.01 to 0.32

0.035

All the covariates were entered simultaneously into the regression models
a

30-item Geriatric Depression score [11

Angiotensin-converting enzyme inhibitors

Instrumental activities of daily living

Table 6 Association (B coefficients, and 95% confidence intervals, CI) between the individual components of the metabolic syndrome and
perceived quality of life, according to the fully adjusted linear regression model, among men
Age \80 (n = 93)

All men (n = 162)

B

95% CI

Abdominal obesity

0.10

-0.04 to 0.24

0.174

Hypertriglyceridemia
Low HDLcholesterol

0.09
0.08

-0.05 to 0.23
-0.08 to 0.25

0.218
0.331

Age C80 (n = 69)

95% CI

0.01

-0.19 to 0.20

0.955

0.18
0.01

-0.03 to 0.39
-0.21 to 0.22

0.099
0.968

95% CI

0.15

-0.01 to 0.31

0.072

0.05
-0.05

-0.10 to 0.21
-0.28 to 0.18

0.497
0.677

High blood pressure

0.01

-0.15 to 0.17

0.903

-0.02

-0.22 to 0.18

0.847

0.06

-0.16 to 0.27

0.598

High fasting blood glucose

0.06

-0.08 to 0.20

0.421

-0.02

-0.20 to 0.17

0.860

-0.01

-0.18 to 0.17

0.958

All the covariates were entered simultaneously into the models

Adjusted for age, education level, alcohol consumption, chronic pulmonary disease, diagnosis of depression, use of corticosteroids and
angiotensin-converting enzyme inhibitors, and instrumental activities of daily living

Table 7 Collinearity diagnostics (variance inflation factors (VIF), and condition index) between the individual components of the metabolic
syndrome and perceived quality of life, among men
All men (n = 162)

Age \80 (n = 93)

Age C80 (n = 69)

VIF

VIF

VIF

Condition index

Condition index

Condition index

Abdominal obesity

1.11

2.10

1.22

2.07

1.10

2.15

Hypertriglyceridemia

1.08

2.45

1.19

2.26

1.04

2.40

Low HDLcholesterol

1.06

2.61

1.08

2.65

1.09

2.72

High blood pressure

1.05

2.65

1.11

2.84

1.04

2.80

High fasting blood glucose

1.03

5.78

1.02

5.15

1.08

7.00

chronic pulmonary disease and depression, use of corticosteroids and angiotensin-converting enzyme inhibitors, and
the instrumental activities of daily living score were all
associated with the HUI3 score at a p \ 0.050 level
(Table 2). After simultaneously adjusting for all these
potential confounders in the summary model (Table 2,

right columns), MetS was not statistically associated with

the HUI3 score in the whole population (B = 0.07, 95%
CI = -0.010.15; p = 0.072). When this fully adjusted
regression model was analyzed after stratifying for sex,
MetS was associated with higher HUI3 score (B = 0.19,
95% CI = 0.060.32; p = 0.006) in men (Table 3); no

123

314

significant associations were found (B = -0.01, 95%

CI = -0.110.09; p = 0.811) in women (Table 3). Also,
logistic regression modeling showed that MetS was associated with an HUI3 score above the median value in men
(OR = 3.34; 95% CI = 1.288.70; p = 0.013) (Table 4).
Analyses of the interaction term indicated that consumption of alcohol did not affect the association between
MetS and a HUI3 score above the median value in the
whole population (p = 0.468), as well as in men
(p = 0.915). Also, use of angiotensin-converting enzyme
inhibitors did not affect such an association in the whole
population (p = 0.586), as well as in men (p = 0.216).
Eventually, use of benzodiazepines did not affect the
association in the whole population (p = 0.635), as well as
in men (p = 0.276).
Also, when the fully adjusted linear regression model
was analyzed in men aged C 80 (Table 5), MetS was statistically associated with the HUI3 score (B = 0.17, 95%
CI = 0.010.32; p = 0.035); no significant association
was found in younger men (B = 0.16, 95% CI =
-0.020.35; p = 0.081). Analysis of the interaction term
confirmed (p = 0.006) that the association of MetS with
HUI3 varied according to age. Also, no significant associations were found in men between the HUI3 score and
any of the single components of MetS, even after stratifying for age below or above 80 years (Table 6). Eventually, collinearity diagnostics indicated that no collinearity
was in men between the HUI3 score and any of the single
components of MetS (Table 7).

Discussion
The results of the present study indicate that MetS is not
associated with worse HRQoL in community-dwelling
elderly; in addition, MetS is associated with significantly
better HRQoL among the oldest male subjects (Table 5).
Due to the aging of populations, the prevalence of MetS
is increasing in Western countries [22]. In keeping with the
hypothesized pathophysiology, which is chiefly based upon
obesity-related insulin resistance with consequent hyperinsulinemia, this condition is associated with increased risk
of diabetes, as well as accelerated atherosclerosis with
ensuing complications, including peripheral vascular disease, stroke, and myocardial infarction [4, 5]. Such complications are known to affect HRQoL, in addition to
survival and functional ability [23]. Knowledge of the
effects of diseases and treatments on patients HRQoL is
crucial for clinical decision making, as well as for resource
allocation, as any interventions in the field of health and
social functioning must have self-assessed, preferencebased quality of life as the main outcome [6]. As stated by
the Panel on Cost-Effectiveness in Health and Medicine

123

Eur J Nutr (2013) 52:307316

conveyed by the US Public Health Service in 1993,

HRQoL should be assessed using comprehensive, generic
classification systems including preferences for a wide
range of health states, based upon time-trade-off measurements, and generating utilities that can be used for
cost-effectiveness calculations [24]. The Health Utilities
Index Mark 3, which has been adopted in the present study,
is among the classification systems approved by the Panel
and is widely used in clinical studies, cost-effectiveness
studies, and in population health surveys. Thus, our results
are of interest for clinicians and decision makers, also
because the cost-effectiveness ratio of preventive interventions for diabetes and cardiovascular disease in subjects
with MetS has been found to depend heavily upon utilities
that patients attribute to their health state [6].
Several studies have investigated the association
between MetS and perceived HRQoL, as well as the possible confounders of such an association, such as gender
[710]. Regarding the first issue, most previous studies,
which involved selected populations of young- to middleaged subjects, have indicated a reduced HRQoL among
subjects with MetS [710]. Indeed, most of these studies
adopted the SF-36 scale to measure HRQoL. This scale is
objective, so it is not conceived, nor suitable, to assess the
impact of pathological conditions (such as MetS) or treatments on self-assessed, preference-based HRQoL. Also,
this scale yields separate scores for general and mental
health; in some cases, MetS can show divergent associations with physical and mental HRQoL scores [25]. The
issue of gender-related differences in the association of
MetS with HRQoL is more complex, because results of
previous studies are often conflicting. Intriguingly, some
studies found that MetS was associated with reduced
HRQoL among women, but not men; in general, among
subjects with MetS, men referred higher HRQoL than
women [8, 9]. Thus, the sex-related differences in the
association between MetS and HRQoL in the present study
add to existing evidence in this field. The determinants of
such gender-related differences are unclear; according to
some studies, it might be related to a stronger association
of MetS with subclinical inflammation in women, as
compared with men [26, 27]. Inflammation, in turn, might
be associated through several pathophysiologic pathways
to decreased HRQoL. However, in the present study, highsensitivity C-reactive protein among participants with
MetS did not differ according to sex. Rather, in our population, MetS has been found to be associated with
depressive symptoms among women, but not in men [28].
This might contribute to the differences observed between
sexes in the association of MetS with HRQoL, as affective
status is included among the dimensions explored by the
HUI3. Indeed, the most remarkable finding of this study,
that was conducted in a whole, unselected older population,

Eur J Nutr (2013) 52:307316

beyond the lack of any significant overall association

between MetS and HRQoL (Table 2), is the increased
HRQoL among the oldest male participants with MetS
(Table 3). These results might be included in the setting of
the so-called reverse epidemiology. This term refers to
repeated observations of established risk factors in the
general population having a paradoxically opposite predictive pattern in selected populations; for instance, obesity, increased serum cholesterol concentration, and higher
blood pressure have been associated with decreased morbidity and mortality in subjects with heart failure or endstage renal disease, as well as in older subjects [29]. In all
these populations, that are characterized by a common
condition of frailty, the presence of obesity, hypertension, or high serum cholesterol levels are thought to reflect
the absence of more powerful risk factors (such as malnutrition) that most commonly, and in a shorter term, might
affect their health status and survival [29]. Therefore,
frailty might represent the common determinant of the lack
of association between MetS with a worse HRQoL, as well
as with the selective association between MetS and better
HRQoL in the present study. Participants in the present
study were markedly older than those enrolled in previous
studies; furthermore, the subgroup of subjects in whom
MetS was associated with better HRQoL was beyond the
life expectancy (around 77 years) of Italian men in 2004.
Thus, the association of MetS with increased health-related
HRQoL among the oldest male, as well as the loss of the
predictive pattern of MetS for decreased HRQoL in the
remaining participants, is likely to reflect the absence of
competitive risk factors, such as hepatic dysfunction,
malnutrition, or cachexia from chronic conditions. In particular, such competitive risk factors might have affected
increased short-term mortality, while MetS might yield its
effects on survival only over a longer time course. Therefore, selective survival might represent the key to understand our, as well as other similar findings of a reverse
epidemiology.
The lack of association between HRQoL and the single
components of MetS (Table 6) might reflect insufficient
statistical power of the single items; indeed, in the multivariable analysis, the B coefficients of the single components seem to diverge, especially after stratifying for age.
Of notice, no collinearity was found among the components of MetS (Table 7). Thus, our results support the
current view that the clinical significance of MetS exceeds
the arithmetic sum of its individual components.
An ineludible limitation of this study is represented by
its cross-sectional design, which does not allow ascertaining any cause-effect relationship. However, independently
of any underlying factors, the neutral or inverse association
on MetS with HRQoL casts doubts on the routine screening
and treatment of MetS beyond the age of 75 and suggests

315

the need of a thorough revision of preventive interventions

in the oldest age segments of populations [29]. In particular, the HUI3 score that we adopted to measure HRQoL
has straight economical implications; according to the
results of the present study, the cost-effectiveness ratio of
interventions aiming at reversing MetS would be less
favorable, or even unacceptable, in older populations.
Acknowledgments This research was partially supported as a
targeted project (SS9.4.2) by the Italian Ministry of Health. The
authors declare that there are no conflicts of interest in this study.

References
1. Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC Jr, Lenfant C,
American Heart Association; National Heart, Lung, and Blood
Institute (2004) Definition of metabolic syndrome: report of the
National Heart, Lung, and Blood Institute/American Heart
Association conference on scientific issues related to definition.
Circulation 109(3):433438
2. Cameron AJ, Shaw JE, Zimmet PZ (2004) The metabolic syndrome: prevalence in worldwide populations. Endocrinol Metab
Clin North Am 33(2):351375
3. Gutierrez DA, Puglisi MJ, Hasty AH (2009) Impact of increased
adipose tissue mass on inflammation, insulin resistance, and
dyslipidemia. Curr Diab Rep 9:2632
4. Qiao Q, Gao W, Zhang L, Nyamdorj R, Tuomilehto J (2007)
Metabolic syndrome and cardiovascular disease. Ann Clin Biochem 44(Pt3):232263
5. Thomas GN, Schooling CM, McGhee SM, Ho SY, Cheung BM,
Wat NM, Janus ED, Lam KS, Lam TH, Hong Kong Cardiovascular Risk Factor Prevalence Study Steering Committee (2007)
Metabolic syndrome increases all-cause and vascular mortality:
the Hong Kong Cardiovascular Risk Factor Study. Clin Endocrinol 66(5):666671
6. Smith KJ, Hsu HE, Roberts MS, Kramer MK, Orchard TJ, Piatt
GA, Seidel MC, Zgibor JC, Bryce CL (2010) Cost-effectiveness
analysis of efforts to reduce risk of type 2 diabetes and cardiovascular disease in southwestern Pennsylvania, 20052007. Prev
Chronic Dis 7(5):A109
7. Tziallas D, Kastanioti C, Kostapanos MS, Skapinakis P, Elisaf
MS, Mavreas V (2011) The impact of the metabolic syndrome on
health-related quality of life: a cross-sectional study in Greece.
Eur J Cardiovasc Nurs, Mar 11
8. Park SS, Yoon YS, Oh SW (2005) Health-related quality of life in
metabolic syndrome: the Korea National Health and Nutrition
Examination Survey 2005. Diabetes Res Clin Pract 91(3):381
388
9. Amiri P, Hosseinpanah F, Rambod M, Montazeri A, Azizi F
(2010) Metabolic syndrome predicts poor health-related quality
of life in women but not in men: Tehran Lipid and Glucose
Study. J Womens Health (Larchmt) 19(6):12011207
10. Sohn YJ, Sohn HS, Kwon JW (2011) Gender differences among
middle-aged Koreans for health-related quality of life related to
metabolic syndrome. Qual Life Res 20(4):583592
11. Ferraroni M, Decarli A, Franceschi S, La Vecchia C, Enard L,
Negri E, Parpinel M, Salvini S (1996) Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol 25(4):
775782
12. Zuccala` G, Onder G, Pedone C, Cesari M, Landi F, Bernabei R,
Cocchi A, Gruppo Italiano di Farmacoepidemiologia nellAnziano
Investigators (2001) Dose-related impact of alcohol consumption on

123

316

13.

14.

15.

16.

17.

18.

19.

20.

Eur J Nutr (2013) 52:307316

cognitive function in advanced age: results of a multicenter survey.
Alcohol Clin Exp Res 25(12):17431748
Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW (1963)
Studies of illness in the aged: the index of ADL: a standardized
measure of biological and psychosocial function. JAMA
185:414419
Lawton MP, Brody EM (1969) Assessment of older: self-maintaining and instrumental activities of daily living. Gerontologist
9(3):179186
Nardi B, De Rosa M, Paciaroni G, Marchesi GF, Bonaiuto S,
Luciani P, Turtu` F, Giannandrea E (1991) Clinical investigation
on depression on a randomized and stratified sample in an elderly
population. Minerva Psichiatr 32(3):135144
Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M,
Leirer VO (19821983) Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr
Res 17(1):3749
Hodkinson HM (1972) Evaluation of a mental test score for
assessment of mental impairment in the elderly. Age Ageing
1(4):233238
Rocca WA, Bonaiuto S, Lippi A, Luciani P, Pistarelli T, Grandinetti A, Cavarzeran F, Amaducci L (1992) Validation of the
Hodkinson abbreviated mental test as a screening instrument for
dementia in an Italian population. Neuroepidemiology 11(46):
288295
PHSHCFA (1980) International classification of diseases, 9th
rev. 1980 Public Health ServiceHealth Care Financing
Administration, Washington, DC
Deyo RA, Cherkin DC, Ciol MA (1992) Adapting a clinical
comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 45(6):613619

123

21. Feeny D, Furlong W, Torrance GW, Goldsmith CH, Zhu Z,

DePauw S, Denton M, Boyle M (2002) Multiattribute and singleattribute utility functions for the health utilities index mark 3
system. Med Care 40(2):113128
22. Eckel RH, Grundy SM, Zimmet PZ (2005) The metabolic syndrome. Lancet 365(9468):14151428
23. Ford ES, Li C (2008) Metabolic syndrome and health-related
quality of life among U.S. adults. Ann Epidemiol 18(3):165171
24. Rich MW, Nease RF (1999) Cost-effectiveness analysis in clinical
practice: the case of heart failure. Arch Intern Med 159(15):16901700
25. Katano S, Nakamura Y, Nakamura A, Suzukamo Y, Murakami
Y, Tanaka T, Okayama A, Miura K, Okamura T, Fukuhara S,
Ueshima H (2001) Relationship between health-related quality of
life and clustering of metabolic syndrome diagnostic components. Qual Life Res [Epub ahead of print]
26. Dupuy AM, Jaussent I, Lacroux A, Durant R, Cristol JP, Delcourt
C (2007) Waist circumference adds to the variance in plasma
C-reactive protein levels in elderly patients with metabolic syndrome. Gerontology 53(6):329339
27. Lai MM, Li CI, Kardia SL, Liu CS, Lin WY, Lee YD, Chang PC,
Lin CC, Li TC (2010) Sex difference in the association of metabolic syndrome with high sensitivity C-reactive protein in a
Taiwanese population. BMC Public Health 10:429
28. Laudisio A, Marzetti E, Pagano F, Pozzi G, Bernabei R, Zuccala`
G (2009) Depressive symptoms and metabolic syndrome: selective association in older women. J Geriatr Psychiatry Neurol
22(4):215222
29. Kalantar-Zadeh K, Block G, Horwich T, Fonarow GC (2004)
Reverse epidemiology of conventional cardiovascular risk factors
in patients with chronic heart failure. J Am Coll Cardiol
43(8):14391444