Prevalence of metabolic syndrome components among the elderly using three different definitions: A cohort study in Finland

Abstract

Objective. Limited data are available on the metabolic syndrome (MetS) and its components in elderly people (aged 70 years and over) at population level in Northern Europe. A study was undertaken to investigate the prevalence of MetS and its components in an aging population by using different definitions. Design, setting, and subjects. A cross-sectional study of 539 inhabitants from Northern Finland (mean age 71.9 years) was conducted to investigate the prevalence of MetS, by using the definitions of MetS by the National Cholesterol Education Panel (NCEP), the modified NCEP (NCEPm), and the International Diabetes Federation (IDF). Main outcome measures. Prevalence of MetS by the NCEP, NCEP modified, and IDF criteria. Results. Overall, the prevalence of MetS was 24.7%, 35.2%, and 37.2% in men, by NCEP, modified NCEP, and IDF-definitions, respectively. In women the corresponding figures were 20.9%, 33.1%, and 47.8%. Hypertension was the most common component in both men (91.8%) and women (89.0%) by the IDF criteria. Glucose abnormalities were particularly prevalent in men (53.2% by NCEP and 78.4% by IDF criteria). Conclusions. The most common component was hypertension in both genders. Lower waist-circumference cut-off points of the IDF criteria led to a higher prevalence of MetS particularly in women. Prevalence of MetS varied significantly when measured by different definitions. Nearly half of older women met the IDF definition of MetS, which was more than twofold when compared with NCEP. Clinical practitioners should be aware of the limitations when using set criteria of MetS, in contrast to identifying the individual cardiovascular risk factors and the accumulation of these.

Key Words::

• Elderly population
• definitions
• IDF
• metabolic syndrome
• prevalence

Data on the prevalence of components of the metabolic syndrome (MetS) among elderly Northern Europeans, particularly as compared by the three currently used different definitions, are limited. The authors observed significant variation in the prevalence of MetS and associated components among an ageing Finnish population, when measured with the different criteria. Furthermore, different definitions identified different individuals as having MetS. Clinical practitioners should therefore be aware of the limitations using set criteria of MetS, in contrast to identifying the individual cardiovascular risk factors and the accumulation of these.

Introduction

The metabolic syndrome (MetS) is a major public health problem worldwide and it is highly prevalent in the adult population [1–4]. MetS is a cluster of cardiovascular risk factors and it is associated with a significantly elevated risk of developing type 2 diabetes and cardiovascular disease (CVD), including in the elderly [5,6].

WHO first defined MetS in 1998 [7], subsequently followed by several other criteria of MetS. The International Diabetes Federation (IDF) has most recently defined MetS in 2005 with central obesity (waist circumference with ethnicity-specific values) as an obligatory item, plus any two of the following: elevated triglycerides levels, blood pressure, fasting plasma glucose, or reduced HDL cholesterol [8, 9]. The IDF definition differs from NCEP [10] and NCEP modified [11] by lower, but obligatory cut-off points of waist circumference.

Limited information has been available on the prevalence of MetS and its components in elderly (aged 70 years and older) Northern European populations, especially as measured by the IDF criteria [12;13]. However, in an adult population the risk of CVD is known to increase as the number of individual components of MetS accumulates [14].

In the decision-making processes of general practitioners, population-level knowledge concerning the prevalence of cardiovascular risk factors is important [15]. Therefore, the aim of the present study was to investigate: (1) the prevalence of the individual components of MetS in an aging population, which forms a significant patient group in general practice, (2) the prevalence of MetS measured by NCEP, modified NCEP, and IDF, and (3) to determine whether these various definitions diagnosed the same individuals with MetS.

Materials and methods

A prospective population-based study was carried out to assess the prevalence of diabetes mellitus (DM) and glucose abnormalities in Oulu, Finland, between 1990 and 2008 [16]. The current study is based on cross-sectional data from the longer follow-up study, which was collected between 2007 and 2008. All living subjects from the original cohort living in Oulu in January 2007 (n =838) were invited to the study, and 64.3% of the subjects (320 women and 219 men) attended. The mean age of participants was 71.9 years (SD ±0.5). The study protocol was approved by the Ethics Committee of Faculty of Medicine, University of Oulu.

Data were collected with postal questionnaires, interviews, clinical examinations, and laboratory investigations. Self-reported data regarding presence of diseases and medication were recorded.

The study protocol included measurements of blood pressure, weight, and height in light clothing. Body mass index (BMI, kg/ m²) was calculated by dividing weight in kg by height in m², and subjects were classified as normal (BMI <25.0), overweight (BMI 25.0–29.9), or obese (BMI ≥30.0). Blood pressure was measured twice with an automatic device (Omron®) by a trained research nurse after 5–10 minutes’ rest in a sitting position. Waist was measured by the research nurse to the nearest centimeter. Venous blood samples were drawn after an overnight fast for plasma glucose and lipids. All plasma samples were analyzed in the laboratory of the University Hospital of Oulu according to a standardized protocol.

MetS was defined according to NCEP [10] modified NCEP [11], and IDF [8,9] criteria. The individual components of different definitions are given in Table I.

Table I. Criteria of the metabolic syndrome according to NCEP, NCEP-modified, and IDF definitions.

NCEP (3 of the following)	NCEP modified (3 of the following)	IDF (waist obligatory +2 other components)
Waist >102 cm (male), >88 cm (female) identifying central obesity	Waist ≥102 cm (male), ≥88 cm (female) identifying central obesity	Waist ≥94 (male) or ≥80 (female) cm or BMI >30.0 kg/m^2
Triglycerides ≥1.7 mmol/L	Triglycerides ≥1.7 mmol/L or specific treatment for this lipid abnormality	Triglycerides ≥1.7 mmol/L or specific treatment for this lipid abnormality
HDL <1.03 mmol/lL (male), <1.29 mmol/l (female)	HDL <1.03 mmol/lL (male), <1.29 mmol/l (female) or treatment for this lipid abnormality	HDL <1.03 (male) or <1.29 (female) mmol/L or specific treatment for this lipid abnormality
Elevated blood pressure ≥130/85 mmHg	Elevated blood pressure ≥130/85 mmHg or antihypertensive medication	Elevated blood pressure: ≥130/85 mmHg or antihypertensive medication
Hyperglycemia: fasting plasma glucose ≥6.1 mmol/L	Hyperglycemia: fasting plasma glucose ≥5.6 mmol/L or treatment for elevated glucose	Hyperglycemia: fasting plasma glucose ≥5.6 mmol/L or previously diagnosed type 2 diabetes

Notes: NCEP = National Cholesterol Education Panel; NCEPm = NCEP modified; IDF = International Diabetes Federation; waist = waist circumference; HDL = high density lipoprotein.

Results

Sixty-nine (12.9%, 95% CI 10.3–16.2) participants (35 male, 34 female) reported having a diagnosis of diabetes. Sixty-three (11.7%, 9.3–15.0) of participants (34 male, 29 female) had previously undiagnosed diabetes, detected by the oral glucose tolerance test (OGTT). Only 75 (34.3%, 28.4–41.5) of the male participants and 161 (50.3%, 45.6–56.9) of female participants had a normal OGTT.

The prevalence of hypertension (use of antihypertensive medication, or blood pressure ≥130/85 mmHg) was 90.2% (95% CI 87.3–92.5) in both sexes (485 participants) according to the modified NCEP and IDF criteria. Although 306 participants (126 male and 180 female) were on antihypertensive medication, 257 (84.0%, 79.4–87.9) of them still had an elevated blood pressure (≥130/85 mmHg). Only 53 (9.8%, 7.5–12.7) of the participants did not use any blood pressure lowering medication and had normal blood pressure.

The prevalence of different components of MetS according to NCEP, modified NCEP, and IDF criteria is shown in Table II. The most common single component was elevated blood pressure, when measured by modified NCEP or IDF criteria (91.8% in men and 89.0% in women). In men the second most common component detected by all three definitions was glucose abnormality, with 53.2% by NCEP and 78.4% by modified NCEP and or IDF. Also in women the second most prevalent single component measured by NCEP and modified NCEP was glucose abnormality, 33.1% and 59.7%, respectively. However, with the IDF criteria, the second most common single component was central obesity, present in 67.3% of women. In men the third most common single component was central obesity as defined by IDF criteria (43.6%). Lipid abnormalities were rarer.

Table II. Prevalence (95% confidence intervals) of single components of metabolic syndrome (MetS) according to the NCEP, modified NCEP and IDF definitions

Definitions and components	Men (n =219)	Women (n =320)	p value
MetS by NCEP	24.7% (19.1–30.9)	20.9% (16.6–25.8)	0.309
MetS by modified NCEP	35.2% (28.9–41.9)	33.1% (28.0–38.6)	0.624
MetS by IDF	37.2% (30.7–43.9)	47.8% (42.2–53.4)	0.015
Elevated blood pressure:
NCEP ≥130/85 mmHg	84.0% (78.5–88.6)	79.0% (74.1–833.3)	0.144
NCEPm ≥130/85 mmHg or treatment	91.8% (87.3–95.1)	89.0% (85.1–92.2)	0.293
IDF ≥130/85 mmHg or treatment	91.8% (87.3–95.1)	89.0% (85.1–92.2)	0.293
Central obesity:
NCEP waist >102 cm (male), >88 cm (female)	21.0% (15.8–27.0)	31.4% (26.3–36.8)	0.008
NCEPm waist ≥102 cm (male), ≥88 cm (female )	23.7% (18.3–29.9)	36.7% (31.4–42.2))	0.002
IDF(obligatory) waist ≥94 cm (male), ≥80 cm (female) or BMI >30	43.6% (36.9–50.4)	67.3% (61.8–72.4)	<0.001
BMI >30	25.8% (20.1–32.2)	31.9% (26.8–37.3)	0.132
Lipids:
NCEP, triglycerides ≥1.7 mmol/L	18.8% (13.9–24.6)	12.6% (9.1–16.7)	0.048
NCEPm, IDF, triglycerides ≥1.7 mmol/L or treatment	18.8% (13.9–24.6)	12.6% (9.1–16.7)	0.048
NCEP, HDL <1.03 mmol/ L(male), <1.29 mmol/ L(female)	8.7% (5.3–13.3)	10.3% (7.2–14.2)	0.531
NCEPm, IDF HDL <1.03 mmol/L(male), <1.29 mmol/ L(female) or treatment	8.7% (5.3–13.3)	10.3% 87.2–14.2)	0.531
Glucose abnormality:
NCEP, FPG ≥6.1 mmol/L or treatment (or previously diagnosed diabetes)	53.2% (46.4–60.0)	33.1% (28.0–38.6)	<0.001
NCEPm, IDF: FPG ≥5.6 mmol/L or treatment for diabetes	78.4% (72.4–83.7)	59.7% (54.1–65.1)	<0.001

Notes: NCEP = National Cholesterol Education Panel; NCEPm = NCEP modified; IDF = International Diabetes Federation; waist = waist circumference; FPG = fasting plasma glucose; BMI = body mass index

Comparison between men and women revealed differences in the prevalence of central obesity, glucose abnormalities, and elevated triglyceride levels. Independent of the definition used, central obesity was more common among women, whilst the latter two were more prevalent among men.

Table III demonstrates the prevalence of MetS in this cohort measured by the three definitions among both sexes. Particularly in women the prevalence of MetS increased from 20.9% to 33.1% through to 47.8% when using NCEP, modified NCEP, and IDF definitions, respectively. A similar pattern in prevalence was seen in men (prevalence of 24.7%, 35.2%, and 37.2%, respectively).

Table III. Prevalence (95% confidence intervals) of the metabolic syndrome (MetS) measured by the NCEP, modified NCEP, and IDF definitions.

MetS	NCEP (%, 95% CI), n	NCEP modified (%, 95% CI), n	IDF (%, 95% CI), n
Male	24.7% (19.1–30.9) 54	35.2% (28.9–41.9) 77	37.2% (30.7–43.9) 80
Female	20.9% (16.6–25.8) 67	33.1% (28.0–38.6) 107	47.8% (42.2–53.4) 153
All	22.4% (18.9–26.0) 121	34.1% (30.1–38.1) 184	43.5% (39.3–47.7) 233

Notes: NCEP = National Cholesterol Education Panel, NCEPm = NCEP modified; IDF = International Diabetes Federation.

There were 19 men in our study who did not fulfill the obligatory waist criteria of IDF but would be diagnosed with MetS by NCEP or modified NCEP criteria. Three such cases were also found among women. In this study population 98 (45%) all men had MetS according to one of these three definitions, but only 42 (19%) of the male population had MetS according to all these definitions. Correspondingly 154 (48%) of the women had MetS by one of these definitions, but only 63 (20%) of the female population according to all three definitions.

Discussion

The most common individual components of MetS in this elderly population were hypertension in both men and women, glucose abnormalities (especially in men), and central obesity (especially in women). Nearly half the women met the IDF definition of MetS, which was over twofold higher compared with the NCEP definition. Also in men the prevalence of MetS by the IDF criteria tended to be higher than that measured by the NCEP criteria.

Depending on the definition, the prevalence of hypertension was 80–90%. It was even higher than that found in the study by Gause-Nilsson et al. in a Swedish elderly population, 68% in men and 58% in women aged 70 years [12]. Gause-Nilsson et al. have suggested that the value of hypertension (≥130/85) may be too low for an elderly population [12]. However, none of the MetS definitions have age-specific values for any individual components. Neither the Finnish Hypertension guidelines [17] nor the European Society of Hypertension Guidelines Committee [18,19] recommendations have different values for the elderly in the management of hypertension. The blood pressure goal remains unclear for elderly patients [18,19]. Nevertheless, a large randomized placebo-controlled study, the Hypertension in the Very Elderly Trial (HYVET), has shown that even patients of 80 years or older do benefit from antihypertensive treatment [20].

Both glucose abnormalities and central obesity were common in our elderly population. It is of note that the Finnish guidelines for type 2 diabetes and hyperglycemia are the same irrespective of age [21]. When using the NCEP definition for glucose abnormalities approximately half the men and one in three women fulfilled the criteria. These figures were higher than in the Swedish elderly population, with glucose abnormalities of 29% in men and 19% in women [12]. Furthermore by the IDF definition (FPG ≥5.6 mmol/l) the prevalence of glucose abnormalities was even higher in our study population, diagnosed in approximately 80% of men and 60% of women.

For abdominal obesity measured by the NCEP definition, the prevalence in our study was somewhat lower than that found in the Swedish study (27% in men and 43% in women) [12]. In our population 43% of men and 67% of women were centrally obese as measured by the IDF criteria. The key difference between the NCEP and IDF definitions is the lower waist-circumference cut-off point in the IDF definition [6,9]. This leads to a higher prevalence of MetS with the IDF definition, especially among women. The IDF definition requires waist circumference as an obligatory measurement and the waist circumference is adjusted for ethnicity [6,7]. The sufficiency of epidemiological data to support the cut-off point for waist circumference has also been discussed [16]. An “obesity paradox” is also common in elderly populations: obesity is associated with increased mortality over the life course; however, obesity observed in old age may be associated with a better prognosis as compared with normal weight [22]. However, recommendations to measure waist circumference rather than the body mass index (BMI) support the importance of abdominal obesity in MetS [23,24].

In a Swedish population the prevalence of MetS as defined by NCEP was 26.2% in men and 19.2% in women [12]. The prevalence of MetS by the NCEP criteria in our cohort is in agreement with the Swedish data. In a Norwegian HUNT study the prevalence of MetS in men and women (aged 70 to 79 years) measured with the IDF criteria was 44.8 and 60.4%, respectively [13]. These prevalences were even higher than in our study.

In our study population of elderly people the prevalence of MetS was high when defined by IDF, but the use of central obesity as an obligatory criterion resulted in the exclusion of subjects with ≥3 of the remaining four criteria, who might also be at increased risk for cardiovascular disease. Approximately half of both the men and women had MetS according to one of these three definitions, but only one-fifth had MetS according to all these definitions. These findings are supported by previous studies showing that different definitions of MetS identify different people as having MetS [25].

Data on MetS using the IDF criteria as a predictor of cardiovascular disease in an elderly population is controversial. Wang et al. have shown that all three definitions under investigation predicted stroke in elderly Finnish subjects during 14-year follow-up [6], and cardiovascular mortality during 13-year follow-up [26]. In another Italian study of elderly population (5632 persons aged 65–84 years), MetS defined by NCEP was associated with stroke or diabetes during four-year follow-up [5]. Conversely, in an Italian study of an elderly population (3038 persons aged 65 to 84 years) the IDF definition of MetS was not associated with a myocardial infarct or stroke at baseline, nor did it predict a myocardial infarct or stroke during three-year follow-up [27].

We conducted a population-based study in an unselected Caucasian population, evaluating the tools for diagnosing MetS in an elderly population, among whom the prevalence of MetS and single components of MetS with different definitions is less known. The study was, however, limited by the relatively small size of the study population. The findings are also limited to a cross-sectional analysis, and we do not yet know the prospective value of the different definitions. The findings from our study limited to an aging Caucasian population, and generalizing them to other populations, must be viewed with caution.

In primary care it is important to find the individuals at high risk of diabetes and cardiovascular disease [28]. The definitions of MetS are possible tools for that, but as we noted in this study and has been noted in others [25,29], there is variation between the performances of different criteria in different populations. Thus there is indeed a need to harmonize the current criteria [24]. In conclusion, the prevalence of hypertension, glucose abnormalities, and central obesity were high in this elderly population. The prevalence of individual components of MetS varied significantly when measured with the different criteria. Clinical practitioners should be aware of the limitations of using set criteria of MetS, in contrast to identifying the individual risk factors for cardiovascular disease.

Ethical approval

The study protocol was approved by Ethics Committee of Faculty of Medicine, University of Oulu.

Declaration of interest

The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.