OUP user menu

The ‘secrets’ of the long livers in Mediterranean islands: the MEDIS study

Eleni Tourlouki, Evangelos Polychronopoulos, Akis Zeimbekis, Nikos Tsakountakis, Vassiliki Bountziouka, Eirini Lioliou, Eftichia Papavenetiou, Anna Polystipioti, George Metallinos, Stefanos Tyrovolas, Efthimios Gotsis, Antonia-Leda Matalas, Christos Lionis, Demosthenes B. Panagiotakos
DOI: http://dx.doi.org/10.1093/eurpub/ckp192 659-664 First published online: 30 November 2009


Background: The aim of the present work was to evaluate various socio-demographic, clinical, lifestyle and psychological characteristics of elderly (>65 years) and very elderly (>90 years) individuals without known cardiovascular disease. Methods: During 2005–7, 1190 elderly (aged >65) men and women (from Cyprus, Mitilini, Samothraki, Cephalonia, Crete, Lemnos, Corfu and Zakynthos) were randomly enrolled. Socio-demographic, clinical, psychological and lifestyle factors were assessed using standard questionnaires and procedures. Results: From all islands, the proportion of males aged 65–80, 80–90 and >90 years was 71.8, 24.8 and 3.4%, respectively. The proportion of women, for the same age categories, was 80.4, 17.9 and 1.7%, respectively. Walking and other activities significantly declined with age (P < 0.001); however, nearly one in five participants over the age of 90 years remained physically active. Current smoking significantly declined in males as age increased (P < 0.001). All participants above the age of 90 years reported sleeping at noon. The proportion of participants living alone differs significantly (P < 0.001) across the three age groups. However, considerably more women live alone (men vs. women living alone: aged 65–80 years 12 vs. 37%; 80–90 years 16 vs. 55%; >90 years 52 vs. 55%). Dietary characteristics of The Mediterranean Islands Study (MEDIS) sample display a favourable adherence to dietary recommendations (Mediterranean diet). Conclusion: A favourable adherence to the Mediterranean diet, mid-day naps and smoking cessation with an increase in age was characteristic of our elderly population. Future research should further evaluate whether the aforementioned characteristics are associated with longevity beyond the average life expectancy.

  • diet
  • elderly
  • long-lived
  • physical activity
  • vital status


It is now evident from various socio-demographic studies that a greater portion of the population survives into old age (over seventh decade of life). However, the factors that promote living after the seventh or eighth decade of life remain unknown. Therefore, a question may arise: what is the ‘formula’ that allows some elders to thwart chronic diseases such as cancer and cardiovascular disease (CVD)? What factors contribute to their improved life expectancy and well-being? The oldest old (≥80 years) in many countries are the fastest growing portion of the population.1 Due to these changes in demography, the United Nations’ Global Population Pyramid is undertaking a shift—from pyramid to cube—as the proportion of children and young adults decline and proportion of elders increase.2 The resultant change in the age distribution of the world’s population is partially attributed to medical advancements of the 21st century, including a decline in infectious and parasitic diseases, a reduction in maternal and infant mortality and improved nutrition.3

Clearly, longevity is a complex attribute determined by factors such as exposure to disease, variability in sleeping patterns, smoking habits, physical activity and diet that have a direct effect, in addition to their indirect emotional and cognitive influence on physiological pathways. Longevity gain marks a significant achievement in human history yet also comes with challenges. The concomitant decline in fertility and mortality rates produces a higher portion of older people, and, thus an increased number of deaths due to non-communicable diseases.4 Thirty percent of the approximated 58 million global deaths from all causes were attributed to CVD—or, otherwise quantified, this proportion is equal to the combined deaths due to infectious disease, nutritional deficiencies and maternal and perinatal conditions.5 Although medical advancements delay mortality, the same cannot be said for preventing disability before death. The World Health Organization reports that 10.3% of total disability-adjusted life years (healthy years of life lost) are lost due to CVD.5 Thus, surviving into old age is only half the battle.

Few epidemiological studies have addressed the question of which factors have the greatest effect on longevity. Trichopoulou et al.6 examined, in a cohort study, the effects of the traditional Mediterranean diet (low in saturated fat, high in monounsaturated fat, high in complex carbohydrates and high in fibre) on health and survival of 182 elderly residents from three Greek villages. They found a one unit increase in diet score (an a priori defined nutritional patterns reflecting the traditional Greek Mediterranean diet) was significantly associated with a 17% reduction in overall mortality. Results of the European Prospective Investigation into Cancer and Nutrition (EPIC) study showed that a modified Mediterranean diet (where unsaturates were substituted for monounsaturates) was associated with increased survival among older people.7 A recently published large longitudinal study showed that longevity was associated with activity, emotional stability and conscientiousness.8 Longevity also seems to be a product of the interaction between physical, social, behavioural and demographic factors.9

The issue of longevity is complex. To begin to understand it one must, at minimum, explore the inter-relationships between factors and to what extent these factors are involved in the cause and prevention of ageing. The present article explores the issue of longevity by presenting a ‘snapshot’ of the various characteristics of this long-lived Mediterranean population. Thus, causal relationships cannot be examined. Under these concepts, it was interesting to evaluate various socio-demographic, clinical, lifestyle and psychological characteristics of old (>65 years) and oldest old (>90 years) individuals without known CVD who lived all their life in various Mediterranean islands and highlight further the concept of longevity on a known healthy European population.


Participants of the study

A random, population-based, multistage sampling method was used to select 553 men (76 ± 7 years) and 637 women (74 ± 7 years), from the Republic of Cyprus (n = 300) and the islands of Mitilini (n = 142), Samothraki (n = 100), Cephalonia (n = 115), Crete (n = 131), Corfu (n = 149), Lemnos (n = 150) and Zakynthos (n = 103) in Greece. Individuals residing in assisted-living centres, as well as those with a clinical history of CVD or cancer were not included in the survey. The participation rate varied according to island, from 75% to 89%. A group of health scientists (physicians, dietitians and nurses) with experience in field investigation collected all the required information, using a quantitative questionnaire and standard procedures.

Clinical, biochemical, socio-demographic and lifestyle characteristics

Study details have been published elsewhere.10 Briefly, clinical, biochemical, socio-demographic and lifestyle characteristics were collected. The evaluated clinical characteristics, as assessed by participants’ physicians, were history and management of diabetes, hypertension, hypercholesterolemia and obesity. Diabetes mellitus (type 2) was determined by fasting plasma glucose tests and was analysed in accordance with the American Diabetes Association diagnostic criteria (i.e. fasting blood glucose levels >125 mg/dl or use of special antidiabetic medication); blood pressure levels ≥140/90 mm Hg or use of antihypertensive medications classified the participants as hypertensive. Fasting blood lipid levels were also recorded and hypercholesterolemia was defined as total serum cholesterol levels >200 mg/dl or the use of lipid-lowering agents. Weight and height were measured to obtain body mass index (BMI) scores (in kilogram per square metre). Obesity was defined as BMI > 29.9 kg/m2, while underweight was BMI < 18.9 kg/m2. Characteristics of the participants were based on a clinical diagnosis made by a physician at hospital or private setting.

Dietary habits were assessed through a semi-quantitative, validated and reproducible food-frequency questionnaire. Intake of various alcoholic beverages was measured in terms of wine glasses adjusted for ethanol intake (e.g. one 100 ml glass of wine was considered to contain 12% ethanol). To evaluate the level of adherence to the Mediterranean diet, the MedDietScore (possible range 0–55) was used where higher values indicate greater adherence to the Mediterranean diet.11 Physical activity was evaluated using the shortened version of the self-reported International Physical Activity Questionnaire (IPAQ).12 Frequency (times per week), duration (minutes per time) and intensity of physical activity during sports, occupation and/or leisure activities were assessed. For the purposes of this work, participants were classified as inactive (i.e. no physical activities during a day) or active. Participants were instructed to report only episodes of activity lasting at least 10 min, since this is the minimum time required to achieve health benefits.

Current smokers were defined as those who smoked at least one cigarette per day or had stopped smoking cigarette during the past 12 months. Former smokers were defined as those who previously smoked, but had not done so for a year or more. Both former and current smokers constituted the group of ever smokers. The remaining participants were defined as rare or non-current smokers. Passive smokers were those exposed to environmental tobacco smoke (ETS), for example, at the workplace, at home or enclosed public areas, for > 30 min/day. Occupational skills were assessed on a scale ranging from 1 to 4. Lower values indicate manual labour, whereas higher values depict labour requiring greater training and theoretical insight. Good or high financial status was defined as >€12 000 per year.

Assessment of depressive symptoms

Symptoms of depression during the past month were assessed using the validated Greek translation of the shortened, self-report Geriatric Depression Scale (GDS).13,14 The GDS questionnaire included ‘yes or no’ items (for details, see Panagiotakos et al.15) where responses were coded one (for answers that indicate depressive symptoms) and zero (for answers that do not indicate depressive symptoms), yielding a possible total score between 0 and 15. Higher values indicate more severe depressive symptomatology.

Statistical analysis

Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as frequencies. Gender-specific comparisons of continuous variables between the three groups (65–80, between 81 and 90 and >90 years) were performed using the Analysis of Variance (for normal distributions) and the Kruskal–Wallis (for skewed distributions). Associations between categorical variables were tested using the Pearson’s chi-square test. All tested hypotheses were two-sided. A P-value <0.05 was considered to be statistically significant. SPSS version 14 software was used for all calculations (SPSS Inc., Chicago, IL, USA).


In table 1, the frequency of males and females of each island, according to age groups (65–80, 80–90, >90 years) is presented. It should be mentioned that during the random sampling procedure the only factor that has been pre-set was the men to women ratio for each island according to the National Statistical Services (varied between 0.95 : 1.05 and 1.01 : 0.99 from island to island). Thus, the frequency of males and females between 80 and 90 years or >90 years is not biased due to sampling procedures.

View this table:
Table 1

Prevalence of age groups in males and females residing in the Mediterranean islands (%)

Island65–80 years80–90 years>90 years

The distribution of various socio-demographic and behavioural characteristics by age group, are given in table 2. Frequency of physical activity varied with age and gender. Men compared with women were more physically active; almost half of the male participants aged 65–90 years were physically active compared with one in four women. Although walking and other activities significantly declined with age (P < 0.001), nearly one in five participants above the age of 90 years was still physically active. The aged-based differences for years of school and occupational skills in men and women were not significant (table 2). Years of school was higher among younger participants and with the exception of women aged 80–90 years, the disparity for occupational skills was slightly higher for men across all age categories. Overall, current smoking significantly declined in males as age increased (P < 0.001). Very few women were current smokers. None of the women over the age of 80 years were current smokers. The number of participants who had ever smoked also declined with age. Men, exceedingly more than women, were ever smokers. For all age categories, nearly one in two men ever smoked, compared with less than 1 in 10 women. Conversely, ETS exposure was more common in women, except in the age category over 90, where 17% of men compared with 0% of women were exposed. Strong associations exist between smoking habits and age categories in men.

View this table:
Table 2

Socio-demographic and behavioural characteristics of the MEDIS study participants according to age category and sex

Males (N = 553)65–80 years (n = 397)80–90 years (n = 137)>90 years (n = 19)P-value
Physical activity (%)4543160.04
Walking and other activities (min./day) (mean ± SD)156 ± 68103 ± 6663 ± 230.001
Years of school, mean ± SD6.8 ± 3.56.0 ± 2.94.7 ± 3.70.008
Occupation skills (1–4) (mean ± SD)1.8 ± 0.91.7 ± 0.81.5 ± 0.70.50
Current smoking (%)3014110.001
Ever smoking (%)6452370.004
Exposed to ETS (%)3517170.004
GDS (0–15) (mean ± SD)7 ± 48 ± 410 ± 30.19
Sleeping at noon (%)86831000.98
Living alone (%)1216520.001
Members of family (mean ± SD)2.2 ± 1.02.0 ± 0.81.6 ± 0.80.002
Good, very good financial status (%)2115210.33
Living in urban areas (%)6351630.04
Age of parents’ death (median for all age groups combined)
Females (N = 637)<80 years (n = 512)80-90 years (n = 114)>90 years (n = 11)P-value
Physical activity (%)2926180.63
Walking and other activities (min/day) (mean ± SD)138 ± 6494 ± 2846 ± 150.001
Years of school (mean ± SD)5.5 ± 3.04.8 ± 2.93.0 ± 2.10.004
Occupation skills (0–4) (mean ± SD)1.7 ± 0.82.7 ± 0.91.3 ± 0.50.25
Current smoking (%)5000.04
Ever smoking (%)8300.08
Exposed to ETS (%)453500.14
GDS (0–15) (mean ± SD)9 ± 48 ± 411 ± 30.19
Sleeping at noon (%)79801000.98
Living alone (%)3755550.001
Members of family (mean ± SD)1.8 ± 0.81.6 ± 0.81.8 ± 1.00.03
Good, very good financial status (%)131400.41
Living in urban areas (%)6456450.16
Age of parents’ death (median for all age groups combined)

Depressive symptomatology as assessed by the GDS scores slightly increased with age, women having slightly higher scores. Nearly all participants, in all age categories reported napping regularly (among men aged 65–80, 80–90 and >90 years, napping rates were 86, 83 and 100%, respectively; among women of the same age categories, rates were 79, 80 and 100%, respectively). Slightly more men than women napped; ultimately, all participants above the age of 90 years reported sleeping at noon. The proportion of participants living alone differ significantly (P < 0.001) across the three age groups, in both men and women. However, considerably more women live alone (men vs. women living alone: 65–80 years 12 vs. 37%; 80–90 years 16 vs. 55%; >90 years 52% vs. 55%). Approximately 50% of all participants lived in urban areas and nearly one in five males reported good or high financial status. More men than women reported a high financial status; 21% of men compared with 0% of women aged >90 years. The age of mothers’ death was slightly higher than the age of fathers’ death for both, men and women aged 65–90 years. The only exception was for participants above the age of 90, where the father outlived the mother (age of father vs. mother, men: 82 vs. 76 years; women 85 vs. 79 years). Men reported a slightly higher number of members in the family than women (table 2).

Dietary characteristics of The Mediterranean Islands Study (MEDIS) sample display a favourable adherence to dietary recommendations (the Mediterranean diet) (table 3). Overall, slight differences exist in the MedDietScore between sex and age categories. Among the men, in the following age categories, 65–80, 80–90 and >90 years, the MedDietScore was 33 ± 4, 33 ± 4 and 34 ± 3, respectively. For the same age categories in women, the MedDietScore was 34 ± 3, 34 ± 3 and 33 ± 3, respectively. Significant differences between sex and intakes (cereals, fruits, vegetables and salads, fish, red meat and products and olive oil) did not exist. At this point, it should be reported that the study population, almost entirely, met dietary recommendations. Cereal intake among men and women differed slightly and met dietary recommendations, with the exception of men in the >90 age category. All participants reported a substantial intake of fruits, meeting and surpassing recommendations (table 3). Frequent consumption of vegetables and salads met recommendations and followed a steady increase with age. Women above the age of 90 years exceeded vegetable recommendations. Dietary recommendations for fish intake are two to three times per week. Younger participants met these recommendations although consumption declined with age. Intakes of red meat and products differed only slightly between men and women; however, all participants, across all age groups exceeded the recommended allowance (table 3). Although a substantial amount of olive oil was consumed by both men and women in all age groups, participants consumed slightly less than the recommendations.

View this table:
Table 3

Dietary characteristics of the MEDIS study according to the age category (mean ± SD)

<80 years80–90 years>90 yearsP-value
    MedDietScore (0–55)33 ± 433 ± 434 ± 30.76
    Consumption of (dietary recommendations)aTimes/weekb
     Cereals (20–30 times)29 ± 1326 ± 1314 ± 140.26
     Fruits (3–4 per day)4.6 ± 1.85.0 ± 1.75.4 ± 1.10.36
     Vegetables and salads (35–60 times)56 ± 3859 ± 4160 ± 370.51
     Fish (2–3 times)2.3 ± 1.22.1 ± 1.31.9 ± 1.10.002
     Red meat and products (0–1 times)2.7 ± 1.02.7 ± 1.32.4 ± 1.10.59
     Olive oil (daily)5.0 ± 1.85.1 ± 1.75.1 ± 1.90.26
    MedDietScore (0–55)34 ± 434 ± 433 ± 30.76
    Consumption of (dietary recommendations)aTimes/weekb
     Cereals (20–30 times)27 ± 1424 ± 1428 ± 180.26
     Fruits (3–4 per day)4.9 ± 1.65.1 ± 1.44.6 ± 2.40.37
     Vegetables and salads (35–60 times)55 ± 3558 ± 4064 ± 330.51
     Fish (2–3 times)2.0 ± 1.11.8 ± 1.11.1 ± 0.70.002
     Red meat and products (0–1 times)2.7 ± 1.02.8 ± 1.02.7 ± 1.10.59
     Olive oil (daily)5.1 ± 1.85.3 ± 1.24.7 ± 1.80.26
  • a: Dietary intakes as per food-based Mediterranean diet recommendations have been included in parenthesis as a point of reference to achieved consumption

  • b: Food consumption refers to times/week unless otherwise noted

An inverse association was observed between anthropometric indices and age category. However, it should be mentioned that these associations were significant, in women, but not in men (table 4). Almost none of the men were underweight; however, 29% aged 65–80 years, 25% aged 80–90 years and 10% aged >90 years were obese. In women, a significant direct association was observed between underweight and age group (P < 0.001). In particular, 0.4% of women aged 65–80 years, 3.7% aged 80–90 years and 10% aged >90 years were underweight. Similarly, a significant association was found between obesity and age category (P < 0.001), where 44, 33 and 10% of women aged 65–80, 80–90 and >90 years, respectively, were obese. Most of the study sample was hypertensive—three out of five participants (table 4). Hypercholesterolemia was prevalent in nearly half the male participants. Moreover, one in five men was diabetic. Strong associations were found regarding hypercholesterolemia and diabetes in women. Hypercholesterolemia was substantially more prevalent in women aged 65–80 and 80–90 years compared with those aged >90 years, 66% and 52% compared with 11% (P < 0.001), respectively. An inverse relationship was also found with the prevalence of diabetes in women where 24, 19 and 9% of women aged 65–80, 80–90 and >90 years, respectively, were diabetic.

View this table:
Table 4

Anthropometric and clinical characteristics of the MEDIS study male and female participants according to age

<80 years80–90 years>90 yearsP-value
    BMI (kg/m2) (mean ± SD)28.1 ± 4.127.4 ± 4.126.6 ± 2.90.09
    Waist circumference (cm) (mean ± SD)105 ± 13103 ± 1199 ± 90.19
    Underweight (%)0.8000.29
    Obesity (%)2925100.16
    Hypertension (%)5971740.03
    Hypercholesterolemia (%)4842390.41
    Diabetes (%)2218260.59
    BMI (kg/m2) (mean ± SD)29.7 ± 5.228.2 ± 5.023.8 ± 4.5<0.001
    Waist circumference (cm) (mean ± SD)103 ± 13104 ± 1293 ± 120.02
    Underweight (%)0.43.710<0.001
    Obesity (%)443310<0.001
    Hypertension (%)7270820.67
    Hypercholesterolemia (%)6652110.001
    Diabetes (%)241990.32


The present work investigated ‘healthy ageing’ of long-lived individuals, who are free of CVD, in order to identify socio-demographic, clinical, lifestyle and psychological characteristics that encourage longevity gain, and, more importantly, quality of life. It is of major interest, nowadays, to study characteristics of people living over the expected life span. Projections for Europe estimate that in 1995, 13.3% of the population was above the age of 65 years, and in 2015, this figure is expected to rise to 16.3%.1 Impressively, in comparison, the present study population included 28.2% and 19.6% of men and women, respectively, who were above the age of 80 years. Data analysis of the MEDIS Study revealed that modifiable risk factors, such as physical activity, diet, smoking cessation and mid-day naps, might depict the ‘secrets’ of the Mediterranean long livers. Similar findings were reported by Knoops et al.16 in the Health Ageing: a Longitudinal study in Europe (HALE) study where lower mortality rates from all causes—coronary heart disease, CVD and cancer—were associated to a Mediterranean diet, moderate alcohol consumption, moderate to high physical activity levels and non-smoking. In addition, one study found regular activity to be associated with reduced overall mortality.17 Rowe and Kahn18 postulate that disability and disease is not an inevitable consequence of ageing—common ageing characteristics are due to lifestyle and other factors that may be age related but are not age dependent. For example, heart disease and type II diabetes may develop in childhood and culminate in later life. This progression occurs over time; however, it can be interceded early in life.

We used a special diet score to estimate the level of adherence to the Mediterranean diet as well as its association with CVD risk factors. Only slight differences in MedDietScore exist in our sample; the scores ranged from 33 to 34 (a fact that may be considered as moderate to good adherence to the traditional Mediterranean diet). Longer survival associated with closer adherence to the Mediterranean diet pattern has been reported by many studies.19–21 High fruit and vegetable consumption, often exceeding dietary recommendations, was a characteristic of this study population. This finding reflects a typical feature of the Mediterranean food culture. For example, green vegetables eaten not only as a salad dish, but also as the main dish of the meal cooked in olive oil. Moreover, wild plants that are frequently collected and utilized as source of food are a widely accepted means of daily living on the Greek islands. Wild greens, more so than red wine or black tea, have exceptionally high flavonoid content—one of the most important categories of antioxidant in the human diet.21 Olive oil, consumed roughly on a daily basis, remained a stable component of the diet across all age groups. Tuck and Hayball22 conclude in their review that olive oil and its major phenolenic constituents are strong antioxidants and good radical scavengers. Consumption of red meat and products surpassed the recommended guidelines (0–1 times/week). It is unclear what effect red meat intake had on clinical variables, though the prevalence of hypercholesterolemia has been related to high intakes of red meat. Frequent consumption of animal products has been directly correlated with mortality from coronary heart disease23 yet, evidence suggests, lean red meat, if consumed in a diet low in saturated fat—the presumed diet of the present population—is associated with reductions in low density lipids cholesterol in both healthy and hypercholesterolemic subjects.24 Dietary recommendations for fish were largely met, predominantly by men. Consumption of fish declined with age and financial status. Fish is often an expensive commodity, and thus its consumption may be limited.

Among women, a significant reduction in BMI and obesity prevalence with an increase in age was observed. BMI and obesity prevalence with an increase in age was also observed in men, although it was not significant. Moreover, waist circumference decreased while the percentage of underweight increased, particularly in women. Ten percent of women aged 90 or above were underweight. The relationship between optimal BMI for longevity remains unclear. Although one cohort found that the leanest women (BMI < 19.0) had the lowest mortality,25 results from a population-based study including Italian elders aged 65–95 years found no relationship between BMI and overall mortality in women.26 This change in weight and height is expected because after the age of 60 years, body weight, on average, tends to decline. Furthermore, Seidell and Visscher27 report that anthropometric data are poor indicators of changes in body composition and fat distribution that accompanies ageing. This may clarify why the relationship between high BMI and mortality is more pronounced in younger populations than older populations. For example, when weight is within the acceptable range or even slightly above, lower rates of mortality are observed in the elderly; whereas, in young or middle-aged populations weight slightly above the acceptable range is associated with higher rates of mortality. Among all participants, the prevalence of hypertension increased with advanced age, whereas an opposite trend was observed for hypercholesterolemia and diabetes. Moreover, almost 9 out of 10 men and all women above the age of 90 years reported no smoking habits. Results from the Cancer Prevention Study II, an ongoing prospective study, suggest that irrespective of age, quitting smoking substantially extends life years.28 Longevity gain for smokers who quit at the age of 65 years was 1.4–2.0 years and 2.7–3.7 years for men and women, respectively. The majority of the participants also took mid-day naps (siesta) almost everyday in the week. In a sample of 23 681 residents from Greece with no history of heart disease, stroke or cancer, Naska et al.29 recently suggested that a mid-day siesta may reduce a person’s risk of death from heart disease, possibly by lowering stress levels. Also, characteristic of the present population was living together with another person (mostly husband and wife, or other family member), which may minimize feelings of loneliness. Work by Arthur,30 on older adult populations in a secondary prevention setting, found a consistent relationship between social support, social isolation and CVD.

The presented findings together with previous works suggest that interplay of genetic, environmental, behavioural and clinical characteristics determine how long an individual lives. This is a widely accepted concept that must be further explored in order to understand how these factors relate and which are most important in shaping longevity. This study cannot fully explain the good health and longevity of these Mediterranean people. However, our observations suggest that modifiable risk factors such as physical activity, diet, smoking cessation and mid-day naps may depict the secret formula of the Mediterranean long livers.


This work was supported by the Hellenic Heart Foundation.

Conflicts of interest: None declared.

Key points

  • Modifiable risk factors such as physical activity, diet, smoking cessation and mid-day naps might depict the secret formula of the Mediterranean long livers.

  • Longevity gain should comprise, in itself, not merely the absence of disease and disability, but the maintenance of high physical and cognitive function, and sustained engagement in social and productive activities.

  • Public health policy and practice must focus on reducing mortality in addition to preventing morbidity in advanced old age to protect the health and livelihood of oldest old.


The authors are, particularly, grateful to the men and women from the islands of Cyprus, Mitilini, Samothraki, Crete, Corfu, Lemnos, Zakynthos and Cephalonia, who participated in and collaborated on this research. The study is funded by research grants from the Hellenic Heart Foundation, and therefore we would also like to thank Prof. Pavlos Toutouzas, Director of the Foundation. The MEDIS study group consisted of M. Tornaritis, N. Papairakleous, S. Papoutsou, A. Polystipioti, M. Economou (field investigators from Cyprus), A. Zeimbekis, K. Gelastopoulou, I. Vlachou (field investigators from Mitilini), C. Lionis, I. Tsiligianni, M. Antonopoulou, N. Tsakountakis, K. Makri (field investigators from Crete), E. Niforatou, V. Alpentzou, M. Voutsadaki, M. Galiatsatos (field investigators from Cephalonia), G. Metallinos, K. Voutsa, E. Lioliou, M. Miheli (field investigators from Corfu) and S. Tyrovolas, G. Pounis, A. Katsarou, E. Papavenetiou, E. Apostolidou, G. Papavassiliou, P. Stravopodis (field investigators from Zakynthos).


View Abstract