Health and Nutritional Status of Elderly Greek Migrants to Melbourne, Australia

Transcription

1 Age and Ageing 1996.: Health and Nutritional Status of Elderly Greek Migrants to, Australia ANTIGONE KOURIS-BLAZOS, MARK L. WAHLQVIST, ANTONIA TRICHOPOULOU, EVANGELOS POLYCHRONOPOULOS, DIMITRIOS TRICHOPOULOS Summary The health (self-reported health conditions) and nutritional status (food and nutrient intake, nutritional biochemistry, anthropometry) of 189 elderly Greeks living in, Australia were described and compared with elderly Greeks living in a rural town in Greece () using a validated health and food frequency questionnaire. was chosen because the traditional diet is maintained by the community and may act as a 'surrogate' measure of diets prevalent in Greece prior to the sample's migration to Australia in the 196s. This enabled identification of dietary trends that may be contributing to the deteriorating health of elderly migrant Greeks. Compared with Greeks, Greeks had significantly greater intakes of animal foods (meat), legumes, protein, margarine, polyunsaturated fats, beer and lower intakes of cereals, carbohydrates, wine and olive oil. The contribution of these dietary differences, as well as the influence of high storage-iron levels, impaired immunity and greater prevalence of obesity and abdominal fatness, to the increasing prevalence of heart disease and cancer (especially amongst women) requires further study. Introduction According to 1982 mortality data [1], Greeks in Australia were deemed the 'second longest lived population in the world', after the Japanese in Hawaii [2] and followed closely by Greeks in Greece [3]. This mortality advantage was mainly due to low rates of coronary heart disease and colonic cancer [4-7]. In contrast, 1989 morbidity data [8] suggest that the health of Greek Australians may have deteriorated since The prevalence of heart disease, hypertension and hypercholesterolaemia was equal to or higher than Australian-born in all age groups. Interestingly, however, the prevalence of cancer remained lower in Greek Australians in all age groups. Studies on small samples of elderly migrants in Australia [9 11] also report that Southern Europeans have worse health and well-being than Australian-born and have a higher prevalence of obesity. It has been suggested that retention of some 'protective' elements of the traditional Greek diet, such as plant food and olive oil, may protect health [2] and reduce mortality rates in elderly age groups [12]. Even though there has been increasing interest in the traditional Greek diet as a healthpromoting nutritional pattern, there have been few attempts to define food and nutrient composition of the Greek diet in Greece [12-17] and Australia [18-2] and identify changes upon migration and with increasing length of stay [6]. Such changes may shed light on possible causes for the deteriorating health of Greek Australians. Most of the studies on Greek Australians provide inadequate and incomplete data on their absolute intake of foods and nutrients and none has included elderly subjects. Elderly people are more likely to adhere to traditional cuisine and thus may provide more useful information on the protective components of such a diet. Greece provides an unusual opportunity because rural regions, such as, still follow a more traditional Greek diet than city regions [17, 21] and can act as a 'surrogate' measure of diets prevalent in the 195-6s, when mass migration to Australia took place. When a longitudinal study is not possible, comparisons of migrant diets with such 'surrogate' measures may provide insight into dietary trends and their potential impact on health status. This paper describes and compares the health status (selfreported health conditions) and nutritional status (food and nutrient intake, nutritional biochemistry, anthropometry) of elderly Greeks living in and. Data on Anglo-Celtic Australians were used for comparative purposes [22]. Methods Sample selection: The study of elderly Greeks in Greece and Australia and of Anglo-Celtic Australians is part of a wider international study of elderly people [22-24]. This study was Downloaded from by guest on 17 November 218

2 i 7 8 A. KOURIS-BLAZOS ET AL. approved by the Monash University ethics committee. The telephone directory has been successfully used to obtain a representative sample of Greek Australians [18] and this method was employed in Australia to recruit elderly Greeks. In countries with low telephone usage, such as Greece, it is preferable to use electoral rolls, and these were therefore used in. All study subjects were interviewed in their homes using an interviewer-administered questionnaire [22, -]. All interviews and measurements were performed by a single observer, fluent in Greek. Residents in psychogeriatric homes were excluded. The total population in in 1988 was about 1, of whom 6.4% were aged 7+. These subjects formed the sampling base. A total of (M 51, F 53) subjects were included in the study (mean age 77 ; 6% 7-79; 4% 8+). The response rate was high at 89% and the sample did not differ in sex and age group distribution from the wider elderly community in. In, the study was conducted between 199 and At the 1986 census in Victoria, persons claimed to be of Greek ancestry, of whom had been born in Greece [29]. Only 2% (2686 persons) were aged 7+. About 3 telephone connections were identified as belonging to Greek families, and these formed the sampling base of the study. A total of 189 (M 94, F95) subjects were included in the study (mean age 78 ; 65% 7-79; 35% 8+). The response rate was 84% and the sample did not differ significantly in sex or age group distribution from the wider elderly Greek community in (1986 Victoria census) except for women aged 7 79 (p =.5). Geographic distribution (percentage living in various suburbs in ) was not significantly different from the wider elderly Greek community in. The mean length of residence in Australia was 3. Health status: This study employed the Multi-level Assessment Instrument [3] which has been recommended as one of the most valid and reliable measures of the health status of elderly people [31]. It includes a 3-item check-list of common self-reported health conditions (see Table I). Health complaints reported were cross-checked by the interviewer with prescribed medications to ensure that complaints were not merely self-perceived but had been diagnosed by a doctor, a consideration particularly relevant for heart disease, cancer and hypertension. This check may not work for conditions that can be treated with diet alone (e.g. diabetes, constipation). Food and nutrient intake: The food frequency questionnaire (FFQ) was aimed at discovering the variety and quantity of foods consumed over the past 12 months. The FFQ from the Australian Polyp Prevention Project [] was adapted to include Greek foods and dishes (5 foods). The portion sizes of foods consumed were elicited in household measures, natural units (e.g. slices of bread) or with reference to food photographs depicting Greek dishes [33]. The FFQ was validated by comparing reported energy intake from the FFQ with minimal energy requirements (MER) calculated from basal metabolic rates (BMR) and the physical activity level (PAL) (MER = BMR x PAL). The MER was calculated for each subject by calculating BMR using the Schofield equations for the 6+ age group [34] and multiplying by an activity factor of 1.55 (assuming a sedentary lifestyle for elderly people) [35, 36]. The mean MERs did not differ significantly from those estimated from the FFQ for men, but did so for the women (p =.5). For about 4% of the women in both sites energy intake estimated from their FFQ was below this MER. More than two-thirds of these subjects were obese, suggesting either under-reporting or self-reported negative energy balance to lose weight. Therefore, when interpreting the food and nutrient intake data, the possibility of under-reporting by obese elderly Greek women should be kept in mind. This also raises questions about the applicability of the Schofield equation to elderly or obese subjects. A combination of the Australian (NUTTAB 1991) and Greek Food Composition Tables [] was used for food and nutrient analyses (Tables II-IV). Macronutrient intake data were also expressed as a percentage of total energy intake (Table III) and the adequacy of micronutrient intakes expressed as a percentage achieving two-thirds of the American Recommended Dietary Intakes (RDI) [37] (Table V). Nutritional biochemistry: Fasting venous blood was collected for the following tests: albumin; total lymphocyte count (TLC); haemoglobin, haematocrit; serum iron, ferritin, ironbinding capacity, iron saturation; vitamin B^ and folate; total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides, LDL:HDL; and glucose. Subjects refusing to give blood were not excluded from the whole study. Blood sampling was preceded by an overnight fast. Both and blood samples were analysed at Monash Medical Centre laboratory in Australia, which complies with WHO Quality Assurance standards. Anthropometry: Anthropometry included height (ht), weight (wt), arm muscle area, and circumferences at the level of the umbilicus (waist) and gluteus maximus (hip). These measurements were used to obtain body mass index (wt/ht 2 ), percentage body fat, lean mass, and waist-to-hip ratio or body fat distribution. To avoid inter-observer variation, these measurements were made by one trained researcher. The measurement procedures adopted match those used in the Euronut-Seneca study [15]. Body fat and lean mass were estimated using the Duerenberg equation [38]; fat free mass (FFM kg) =.2* ht +.395* wt + 8.4* sex -.4* age 23.6 (height in cm, weight in kg, sex = 1 for men and for women, age in ). Arm muscle area (cm ) was calculated from arm circumference and triceps skin-folds [39]. Data analysis: Subjects were recruited from the age groups 7-79 and 8+. Even though the number of observations was small for some variables, these age groups were analysed separately as they appeared to show different trends. The Statistical Analysis System (SAS, 1993) was used for analyses. Non-parametric statistics (Wilcoxon rank sum test; continuous variables) and x (discrete variables) were used to test the significance of differences between sex, age group and centre (location). Results Results are specific to the community studied and may not be extrapolated to the wider elderly Greek population in Greece and Australia. Health status: The ten most common self-reported health conditions are presented in Table I. For conditions that require prescribed medications for treatment, e.g. hypertension, heart disease, there was agreement between the proportions reporting the complaint and the proportions taking the prescribed medication. More than 9% of subjects reported at least one health condition. Significantly more women (7%) than men (55%) reported three health complaints. The most common health problem was hypertension (45%), followed by arthritis (3%). The proportions of women (mainly aged 8+) reporting heart problems and cancer (4%, % respectively) were significantly greater than in the men (23%, 4%) and Downloaded from by guest on 17 November 218

3 HEALTH AND NUTRITIONAL STATUS OF ELDERLY GREEK MIGRANTS 179 Table I. The ten most common self-reported health conditions (%) and prescribed medications in elderly Greeks living in, Greece and, Australia Men(n) 1. Hypertension 2. Arthritis 3. Impaired hearing 4. Impaired vision 5. Heart trouble 6. Diabetes 7. Constipation 8. Ulcer 9. Stroke 1. Cancer (excludes skin) Women (n) 1. Hypertension 2. Arthritis 3. Impaired hearing 4. Impaired vision 5. Heart trouble 6. Diabetes 7. Constipation 8. Ulcer 9. Stroke 1. Cancer (excludes skin) * 18.8 e k 12.9"" 12.9 s & k 3.2 Mx* (4.6) () () (6.2) () () (38.7) (12.9) (9.6) (12.9) () (3.2) jc b " 5. 4 ' b Pairs of letters indicate significant differences, x 2 P <.5): a,b,c or d within centres between sex for a given age group; e,f,g or h within centres between age groups for a given sex; i j,k or 1 between centres for a given age group and sex. Mx = Medication (prescribed). women (%, 4%). The prevalence of diabetes was high in both centres (M 1%, F 2%). Less common health complaints included stroke, ulcers and cancer. VEGETABLES MILK MEAT SUGAR PRODUCTS LEGUMES BEER MARGARINE CEREALS WINE OLIVE OIL FRUIT FISH CHEESE YOGHURT EGGS SOB (%) Mx (41.) () (2.9) () (1.5) (5.2) (45.4) (9.) (13.6) (18.1) (4.5) () C f j f 3.4 f * 15.3 h O.5 h k k 3. (%) Mx (42.4) (15.1) (24.2) (12) (3.) (7.5) (47.4) (42.3) (37.2) (1.1) (5.) (1.1) d fj f CO"" " 44.4 h 13.6' 5." ' d (%) Mx (46.3) (17.8) (35.7) (7.1) (3.5) (3.5) (52.7) (47.2) (44.3) (13.8) (13.8) (5.5) Food intake: Compared with Greeks, Greeks had a significantly greater mean daily intake of total foods (5g vs. 12g); animal foods (5g vs. 4 g), and plant foods (95 g vs. 8g). Differences in GREATER INTAKE IN MELB. LOWER INTAKE IN MELB. SIMILAR INTAKE IN MBLB. Downloaded from by guest on 17 November MELB. 2 g/day SPATA 3 4 Figure 1. Differences in food group intake between elderly Greeks in and (p <.5).

4 i8o A. KOURIS-BLAZOS ET AL. Table II. Comparison of daily food consumption of and Greek elderly men and women by quantitative food frequency questionnaire: percentiles with (mean SD) (g/day)* (g/day)* Men{n) Percentiles % Total food Animal food % total food Meat/poultry Fish Milk Yoghurt Cheese Eggs Plant food % total food Vegetables Legumes Cereals Fruit Plant/animal Fats Olive oil Sugars Alcohol Beer Wine Women (n) Percentiles % Total food Animal food % total food Meat/poultry Fish { ") {389 2') 27 { 1') {98 51") (6 4) 69 (11 156) 12 (31 41) 3 43 (44 2) 4 (11 15) ( "') (72 1') (378 3'") 3 48 (6 42") (4 1") ( ) ( ') (34 12) (34 12') (37 39') 2 (8 9') (6 27') 2 (2 5 l ) (9 315*) (336 ) 29 ( 11) (71 4*) 17 5 (48 42) (1212 b >) (4 193') 31 (35 13) 63 9 (186 bl ) 52 (49 3) 12 (97 78') 19 (35 4) (47 29) 8 (17) (781 2") 6 64 (65 13) (274 3") (44 26') ( ) (189 92) ( ) 35 (35 13 b >) 35 (3412 bi ) 1 (68 19) 1 (27243 b ) ( 1) 1 ( ') (18349 bl ) (385 15) (36 9) 63 (86 34 b ) (4 37) (3 421 ) ( ") 31 ( 8') (158 62") (63 41) 7 (5 158 f ) 11 (26 4) (47 31) 4 (11 18) (183 7"') (68 8') ( *') (93 59') (264 98") (24 2) (2.6 1.') 17 (313 c ) 15 2 (21 15") (119 2') 5 (11 C ) (46 92') (76 123') (1295 *) ( C ) 31 ( 11) 73 (19 52*) (49 37) ( d> ) 42 5 ( *') 34 (38 13) (965 di ) 43 (48 35) 1 2 (27312 dt ') 11 (27 38) 15 4 (36 22) 3 6 ( 3) ( d!i ) (62 13) ( r> ) (7848 di ) (268 1) (2 ) ( ) 2 (27 11') 8 15 (18 13') 41 (11152) (124 8 d ) (5218 d ) (68 112') ( dl ) ( d ) (34 9) (99 53 d ) 3 (41 34) Downloaded from by guest on 17 November 218

5 HEALTH AND NUTRITIONAL STATUS OF ELDERLY GREEK MIGRANTS Table II. (continued) (g/day)» (g/day) # Percentiles % Milk Yoghurt Cheese Eggs Plant food % total food Vegetables Legumes Cereals Fruit Plant/animal Fats Olive oil Sugars Alcohol 5 86 (96 9) 5 19 ( 31) (43 45) 2 (677) (7 3*) (68 11) (297 4*) (38 33*) 6 2 (227 99') 97 3 (776 95) ( ) (J72 k ) (29 77 k ) 1 17 (45 61) ( 45') (77 59') 7 38 (42 44) (38 37) 3 (1) (6955 l ) (64 9) ( ') (48 29 ] ) ( ) 69 1 (3 129) (2..8) 15 ( 11 b ) 15 ( 11 bl ) 1 15 (56 13) (4577 b ) (76 99 h ) (24 33) 48 (49J7 h ) 2 6 (97J) (885 1*) (6S77) 5 39 (4 5***) (77 59^) 8 23 ( 76 hc ) (2 122) ( ) 15 ( 9*) 3 15 ( 77"*) 3 15 (#7727) (48 12") Pairs of letters indicate significant differences in mean intakes, Wilcoxon p <.5; a,b,c or d within centres between sex for a given age group; e,f,g or h within centres between age groups for a give sex; i j,k or 1 between centres for a given age group and sex. Includes non-consumers. PROTEIN %E SIMPLE CARBO%E POLYUNSAT.FATS %E FIBRE g/day TOTAL CARBO. %E COMPLEX CARBO %E - j GREATER INTAKE IN MELB.. i i.. i.. i^^^^ ("IWCn IMTAtfF IN MFI B ( ) 8 (29 44) 31 (297 7 h ) 4 8 (77 73) (8368 M ) (65 9) (6 136 M ) (55 4<S dhj ) 1 2 (24799 h ) (77S) ( ) ( 11) 8 15 (1813 X ) 7 31 (83 131) (3355 d ) Downloaded from by guest on 17 November 218 TOTAL FATS %B MONO FATS %E SAT FATS %E ALCOHOL %E Hsa i : SIMILAR INTAKE IN MELB. i 1 m MELB i i. i SPATA Figure 2. Differences in macronutrient intake of elderly Greeks in and (p <.5).

6 182 A. KOURIS-BLAZOS ET AL. food group intake are summarized in Figure 1 and Table II. men aged 7-79 obtained a greater proportion of their total food intake from animal foods (35%) and a smaller proportion from plant foods (65%) compared with men (31%, 69% respectively). This was also reflected in the significantly lower plant to animal food ratio in (2.6) compared with (3.2). In contrast, women in and had similar ratios (33%, 67% respectively). For cereal consumption, centre differences were only seen for men aged 7 79 ( 3g/day; 266g/day). Nutrient intake: The and samples had a similar intake of total calories (M 23 kcal, F 18 kcal), the higher total food intake of Greeks being offset by their lower intake of olive oil, fats and alcohol. The differences in macronutrient intake are summarized in Figure 2 and Table III. More than 9% of subjects were not achieving the recommended energy intake from complex (4 5%) and total carbohydrates (5-6%). In contrast, 3-5% of the subjects were consuming more than 15% of their energy intake from refined carbohydrates compared with less than 1% of the sample. Only 8% of men and none of the women had fibre intakes above the recommended 3g/day compared with 21% of women and 35% of men. More than 95% of subjects had fat intakes above the Table III. Comparison of daily macronutrient consumption of and Greek elderly men and women (including non-consumers) by quantitative food frequency questionnaire No. of subjects Energy (kcal) Carbohydrate Complex Refined Fibre Protein Fat Saturated Mono-unsat. Polunsat. Cholesterol (mg) Alcohol Men y " ' ' ! " ' b j ej j J b C ' ' ' \& 19. 3' C C d ! C j > J d > d Women ' b k k k k " k l k *.5 1' I b c " k k k C C k * J C d h ' ' d Downloaded from by guest on 17 November 218 Values are means standard deviations. Pairs of letters indicate significant differences, Wilcoxon p <.5: a,b,c or d within centres between sexes for a given age group; e,f,g or h within centres between age groups for a given sex; i j,k or 1 between centres for a given age group and sex.

7 HEALTH AND NUTRITIONAL STATUS OF ELDERLY GREEK MIGRANTS 183 Table IV. Comparison of daily mineral and vitamin consumption of and Greek elderly men and women by quantitative food frequency questionnaire Men No. of subjects Sodium* (mg) Potassium (mg) Calcium (mg) Phosphorus (mg) Magnesium (mg) Iron (mg) Zinc (mg) Vitamin A RE(Mg) Retinol (jig) Carotene (/xg) Thiamin (mg) Riboflavin (mg) Niacin (mg) Vitamin C (mg) Women No. of subjects Sodium* (mg) Potassium (mg) Calcium (mg) Phosphorus (mg) Magnesium (mg) Iron (mg) Zinc (mg) Vitamin A RE (jig) Retinol (jj.g) Carotene (jxg) Thiamin (mg) Riboflavin (mg) Niacin (mg) Vitamin C (mg) l' ' 63! * '.9.3 ei 1.3.4' ' ! k O3 k k k k k.7.2 k 1..3 k k 73.6 k j J J.7.2^ J.5 9 J j ' ' ' '.7.2 bl 1.1.4' ' ' i 34 98' ' 1 51 a di 1.3.5' 1.8.6' ' " h 377 k k k k * dhk 1..3 k k k k Values relate only to non-discretionary salt (i.e. does not include salt added to cooking/at table). Values are means standard deviation. Pairs of letters indicate significant differences, Wilcoxon p <.5: a,b,c or d within centres between sexes for a given age group; e,f,g or h within centres between age groups for a given sex; ij,k or 1 between centres for a given age group and sex. recommended maximum of 3% calories from fat. Comparing the absolute intake of micronutrients in the two samples, subjects had a significantly greater intake of potassium (3 mg vs. 22 mg); phosphorus (13mg vs. 1 mg); magnesium (3 mg vs. 2 mg); zinc (mg vs. 13 mg); vitamin A (9RE vs. 6RE); thiamin (1.2mg vs. mg); riboflavin (1.6 mg vs. 1.2 mg); niacin (37 mg vs. 26 mg) and vitamin C (15mg vs. mg) and a similar intake of calcium (M 7 mg, F 6 mg); iron (2 mg) and non-discretionary sodium (M 22 mg F 17mg) (Table IV). A greater proportion of subjects had potassium intakes below 18 mg (M 15%, F 45%) j j j J 1.2O.3 J j J J h * ' 246 7' ' w.9.3' 1.4.4' ' ' compared with the sample (M 2%, F 11%) (Table V). Nutritional biochemistry: When interpreting the blood results of this study, one should keep in mind that the average participation rate for blood testing was 52% giving a total of 67 (M 4, F 27) subjects in and 17 (M 6, F 47) subjects in. The results reported, therefore, are not truly representative of all elderly Greeks living in and (Table VI). All subjects had serum albumin values above 35 g/1, except among women aged 8+, 7% of whom had values of 34 g/1. men aged 7-79 had a significantly greater TLC than Downloaded from by guest on 17 November 218

8 184 A. KOURIS-BLAZOS ET AL. Table V. Percentages consuming less than two-thirds of the United States Recommended Daily Intake for minerals and vitamins in and Men Women No. of subjects Sodium* (mg) Potassium (mg) Calcium (mg) Phosphorus (mg) Magnesium (mg) Iron (mg) Zinc (mg) Vitamin A (RE/ig) Thiamin (mg) Riboflavin (mg) Niacin (mg) Vitamin C (mg) ' Non-discretionary only men aged women aged 7 79 and men aged 8+ also had a greater TLC than subjects. The prevalence of low TLC (< 15 mm ) was significantly greater in (M %, F 15%) than (7%). Iron deficiency anaemia (TIBC >55 /umol/1, transferrin saturation <15%) was found in six subjects. The prevalence of iron deficiency anaemia was greater in (M 2%, F 33%) than (M 13%, F 17%). Storage-iron levels of >3/ig/l plasma ferritin for men and >5^g/l for women are considered high in elderly people. Significantly more subjects had values in this range (M 19%, F 22%) than did the subjects (M 3%, F 4%). Plasma folate concentrations were satisfactory in all subjects. The mean levels for plasma B ]2 were above the criterion of 111 pmol, used to define a high risk of deficiency, but 1% of the men had levels <111 pmol. The majority of subjects (M 7%, F 78%) and subjects (M 66%, F 67%) had serum total cholesterol levels above 5.5 mmol/l. A greater proportion of men (%) than women (7%) had serum triglycerides above 2. mmol/l; the proportions in were 1% of men and 13% of women. A greater proportion of subjects (M 95%, F 96%) had HDL cholesterol levels above 1 mmol/l (men) or 1.2 mmol/l (women) than did the subjects (M 56%, F 7%). A smaller proportion of men (mainly aged 7-79) had LDL cholesterol levels above 4 mmol/l (45%) compared with women (67%) and subjects (M 56%, F 57%). The prevalence of diabetes (fasting glucose >7.7 mmol/l) [39] was greater in women in (2%) than in (8%). Anthropometry: In, anthropometry was performed only on those subjects who had their blood sampled (M 4, F 27), therefore results may not be representative of all elderly Greeks living in. In contrast, all subjects in the sample underwent anthropometry (Table VII). A significantly greater proportion of subjects were obese (F 45%, M 3%) compared with the sample (<3%). The women tended to be more overweight than the men and to have higher waist-hip ratios (WHR). Almost all subjects had WHR above the recommended level. Less than 1% of (mainly women aged 8+) and subjects (mainly men aged 8+) were at risk of protein energy malnutrition reflected by their low BMIs and arm muscle area values. The mean percentages of body fat and lean body mass were similar in the centres (body fat M 33%, F 49%; lean body mass M 48 kg, F 33 kg). Discussion Southern Europeans (SEA) living in Europe and Australia have been reported to have low morbidity and mortality rates from heart disease and cancer, particularly colon and breast cancer [1-3, 4, 42-44]. Mortality and morbidity from these diseases especially coronary heart disease have recently been shown to be increasing in a steady and alarming way amongst Greeks in Greece and in Australia [3, 15, 43, 46, 47]. In Australia, the prevalence of heart disease amongst SEA has reached the high levels found amongst the Australian-born in all age groups [8], whereas they continue to enjoy lower rates of cancer. SEA had about half the prevalence of cancer seen in the Australian born, except in the older age groups where the prevalence was found to be similar amongst women although lower for SEA men [8]. The current study also reflects these trends. and men Downloaded from by guest on 17 November 218

9 HEALTH AND NUTRITIONAL STATUS OF ELDERLY GREEK MIGRANTS.85 Table VI. Comparison of nutritional biochemistn' (fasting) of and samples (mean SD) Men (n) Serum albumin (g/1) Total lymphocyte count (mm ) % lymphocytes (tlc/wbc x 1) Haemoglobin (g/dl) Haematocrit (%) Plasma iron (jimol/1) Total iron binding (/imol/1) Transferrin saturation (%) Plasma ferritin (jxg/l) Plasma folate (nmol/1) Plasma B 12 (pmol/1) Serum cholesterol (mmol/1) Serum triglycerides (mmol/1) Serum HDL-cholesterol (mmol/1) Serum LDL-cholesterol (mmol/1) LDL/HDL Fasting plasma glucose (mmol/1) Women (n) Serum albumin (g/1) Total lymphocyte count (mm ) % lymphocytes (tlc/wbc x 1) Haemoglobin (g/dl) Haematocrit (%) Plasma iron (/imol/1) Total iron binding (/jmol/1) Transferrin saturation (%) Plasma ferritin (pg/1) Plasma folate (nmol/1) Plasma Bi 2 (pmol/1) Serum cholesterol (mmol/1) Serum triglycerides (mmol/1) Serum HDL-cholesterol (mmol/1) Serum LDL-cholesterol (mmol/1) LD1/HDL Fasting plasma glucose (mmol/1) ' 33 12'.2 l ai " ! " " ! k ak ^ * " " Pairs of letters indicate significant differences, Wilcoxon p <.5: a,b,c or d within centres between sexes for a given age group; e,f,g or h within centres between age groups for a given sex; i j,k or 1 between centres for a given age group and sex. reported a similarly high prevalence of heart trouble (23%) comparable with the prevalence reported in the elderly Anglo-Celtic Australian study [22] and elderly Greek subjects in the Euronut-Seneca study [15]. In contrast with the men, Greek women in (mainly aged 8+) had a significantly greater prevalence of heart disease (4%) than the men, the women (%) and the women in the elderly Anglo- Celtic Australian study [22]. and men also reported a similarly low prevalence of cancer (4%, excluding skin cancer) which was significantly lower than the prevalence reported in the elderly Anglo- Celtic Australian study (>7%, excluding skin cancer) but comparable with the prevalence reported in elderly '.3 2 b b b J b b ^ b b 2.9 6' bl ' bl ' b C C.9 1 C C C ! f a l d ck 22 c chk 13.5 l ck C c c c d j d d j f dl h dl d ' ' ' Greeks in the Euronut study [15]. women (mainly aged 8+) had a greater prevalence of cancer than the men and the women and the women in the Anglo-Celtic Australian study. This suggests that women are losing their protection against heart disease, and also cancer, at a faster rate than the men and the women. The traditional Greek diet can be defined as vegetarian, foods from plant sources forming the core of the diet, while foods from animal sources provide the fringe of the diet (plant to animal food ratio : ) [17, 48, 49]. This dietary pattern, prevalent prior to the 196s, has been associated with a lower prevalence of heart disease, colonic cancer [5, 6, 42] and reduced risk Downloaded from by guest on 17 November 218

10 186 A. KOURIS-BLAZOS ET AL. Table VII. Comparison of anthropometnc measurements of and samples (mean SD) Men(n) Height (cm) Weight (kg) BMI (kg/height m 2 ) %<2 % % % 5=3 Estimated body fat (%) Estimated lean mass (kg)* Arm muscle area (cm 2 ) Umbilical circumference Maximal gluteal circumference Umbilical/maximal gluteal % $=.9 Women (n) Height (cm) Weight (kg) BMI %<2 % % % >3 Estimated body fat (%) Estimated lean mass (kg)* Arm muscle area (cm 2 ) Umbilical circumference Maximal gluteal circumference Umbilical/maximal gluteal % 3s *.7 * " 5* 51 6" " " " * * 8k 47 3" 34 5" 2.81*«k * b ^' 4 b 46 5** e b bj b ^ 492 b 3 b E b b 1. Pairs of letters indicate significant differences, x* (%) or Wilcoxon (score), p <.5: a,b,c or d within centres between sexes for a given age group; e,f,g or h within centres between age groups for a given sex; ij,k or 1 between centres for a given age group and sex. Duerenberg equation (see Methods). of death in later life [12], and is now regarded as a prudent diet [17, 49, 5]. The traditional Greek diet translates into the following macronutrient proportions; low-moderate protein intake (<15% of energy intake); high total fat intake (4-45%); high monounsaturated fat intake (2-22%); low saturated fat intake (1-12%); low polyunsaturated fat intake (<5%); moderate carbohydrate intake (38-44%; complex %; refined 12%) and low-moderate alcohol intake (<5%) [, 17]. In recent, Greeks in Greece [21] and abroad [2, 4, 5] have been moving away from the traditional dietary pattern, towards a more 'affluent' diet, characterized by a proportionately higher animal food intake (especially marked in migrants) and lower plant food intake. Since was chosen for its 'traditionality' it was not surprising that the plant to animal food ratio was significantly greater in (69:31) than in (65:35), with the latter approaching the lower ratio found C 76.3 ll c C * 334 C 515 rf rf C c C 68.8ll ch ch chk 483 C 355 ch 49.7ll chk c c d 72.3 ll d dj 344 d 47 6 df df j d 61.4 dh h dhj 5O3 d 36 dh dh amongst elderly Anglo-Celtic Australians (59:41) [22]. Energy intakes from the macronutrients were in the realm of the traditional Greek diet for both locations, except that subjects had lower intakes of complex carbohydrates (37%) and greater intakes of refined carbohydrates (%; mainly from fruit juice), protein (19%; mainly from meat) and polyunsaturated fat (6%; mainly from margarine), similar to levels found in Anglo-Celtic Australians [22, 49]. The greater intakes of meat (13 g) and protein (19% energy) by Greeks are approaching the intakes of elderly Anglo-Celtic Australians (g/ day; 19% energy) [22]. Similar findings have been reported in other studies [4, 6]. Even though meat intake was significantly lower in (9g/day), this was still greater than intakes reported in Greece in the 196s (35 g/day) [44]. In contrast, fish consumption has remained high in both and (5 g/day) since the 196s (4 g/day) and has not dropped to the Downloaded from by guest on 17 November 218

11 HEALTH AND NUTRITIONAL STATUS OF ELDERLY GREEK MIGRANTS 187 lower levels reported in Anglo-Celtic Australians (15 g/ day) [22, 51]. National authorities are now inclined to recommend decreased consumption of animal protein in preference for plant proteins, which have been associated with reduced rates of heart disease and colonic cancer [52, 53]. Continued high intake of animal proteins in Greeks may be contributing to the changing prevalence of these diseases. Greek elderly subjects (especially men) have also maintained the high vegetable (4 g/day) and legume (7 g/day) intake of the 196s [51] but have markedly decreased their consumption of cereals ( g/day), complex carbohydrates (22% energy intake), and fruit (2 g/day). Elderly Anglo-Celtic Australians have high plant food intakes similar to Greeks, but the types of plant foods consumed are significantly different namely a lower intake of legumes (2 g/day, comprised mainly of peas) and cereals (comprised mainly of bread and breakfast cereals). Certain legumes and cereals have been shown to lower cholesterol and to be potentially protective against colonic cancer [53, 54]. Greeks may have continued protection against colon cancer relative to Anglo-Celtic Australians owing to their higher intakes of specific legumes (e.g. dried beans, chick-peas) and cereals (rice, pasta). The high intake of total fat by elderly Greeks (4 42% energy intake) is also in agreement with other studies in Greece [15-17]. Case-control studies from Greece have not shown adverse effects of the currently high fat intake (4-45%) on risk of cancer, coronary heart disease or obesity. Researchers in Greece argue that the current evidence does not justify reduction of fat intake in the Greek diet if derived mainly from mono-unsaturated fats [, 41, 43], particularly since olive oil facilitates increased consumption of plant foods [12] and has been associated with reduced rates of heart disease [54]. Olive oil consumption has been reported to have decreased markedly in Greece since the 196s (from about 6 g/day [51, 55] and to have decreased even further in migrant Greeks [5, 6]. Much of this olive oil has been replaced by margarines and other polyunsaturated oils, which have recently been linked to heart disease and cancer [53]. elderly subjects consumed only olive oil (3 g/day) which was an important marker of adherence to the traditional Greek diet. In the Euronut study, Greek elderly subjects in a rural village reported eating about 3 g/ day [12]. In contrast, elderly consumed significantly less olive oil (18 g/day) and had introduced margarine into their diets (3 g/day), but not to the levels found in elderly Anglo-Celtic Australians (15 g/day) [22]. Red wine has also been associated with reduced incidence and mortality from coronary heart disease [56], whereas beer has been epidemiologically linked with rectal cancer [57] and heart disease [5, 58]. About 9% of the alcohol consumed by men was from locally produced wine (17 g/day; 6% red, 4% white). In, only 58% of the alcohol consumed by men was from wine (7 g/day); 37% was as beer (5 g/ day) and 5% as spirits. The reduced intake of red wine and olive oil and the increased intake of margarine and beer may be playing a significant role in the changing health profile of migrant Greeks. With respect to micronutrient intake, the diets of Greeks were more nutritionally adequate than Greeks, owing to a significantly greater intake of meat, vegetables, legumes and milk. The nutrients at greatest risk of deficiency included thiamin, magnesium, and vitamin A, with more than 5% of elderly subjects failing to achieve two-thirds of the RDI compared with less than 2% of subjects. The risk of deficiency was lower for calcium, followed by zinc and lastly riboflavin, in both study sites. Vitamin C, niacin, iron, potassium (except for women) and phosphorus were least likely to be inadequately consumed, especially in. Published studies of elderly people have also reported similarly low intakes of these nutrients [15, 59, 6]. Measurements of body fatness and nutritional biochemistry revealed some important differences between the two locations. Greeks (particularly the women) were at greater risk of disease than Greeks because of the greater prevalence of morbid obesity, centrally distributed fat [61], unfavourable blood lipid profiles and impaired immunity [62]. The immune function of Greeks appears to be dropping to the lower levels found in Anglo-Celtic Australians [22]. High levels of storage iron have been associated with impaired immunity, as well as heart disease and colon cancer [63]. High storage iron levels were more prevalent in (2%) than (3%), approaching the high levels found in Anglo-Celtic Australians (>2%). The extent to which iron status contributes to the deteriorating health and immune function of elderly Greeks needs further investigation. In conclusion, there is evidence that elderly Greeks (mainly men aged 7 79) are 'healthier' than elderly Greeks, especially with respect to blood lipid profiles and immune function. Compared with Greeks, Greeks had significantly greater intakes of animal foods (meat), legumes, protein, margarine, polyunsaturated fats, and beer, and lower intakes of cereals, carbohydrates, wine and olive oil. The contribution of these dietary differences, as well as the influence of high storage iron levels, impaired immunity and greater prevalence of obesity and abdominal fatness, to the increasing prevalence of heart disease and cancer, especially amongst women, merits further study. Acknowledgements This study was funded in part by the Sandoz Foundation for Gerontological Research. The statistical assistance of Dr Bridget Hsu-Hage and the co-operation of the staff at the Community Health Centre are gratefully acknowledged. References 1. Young C. Selection and turvival: immigrant mortality in Australia. Studies in Adult Migrant Education. Depart- Downloaded from by guest on 17 November 218

12 188 A. KOURIS-BLAZOS ET AL. ment of Immigration and Ethnic Affairs. Australian Government Printing Services, Canberra, Australia, Powles JW. Best of both worlds? Attempting to explain the persisting low mortality of Greek migrants to Australia. In: Caldwell JC, ed. Health transition: cultural, social and behavioural determinants of health: what is the evidence? Proceedings of a workshop, Canberra May Australian National University, Canberra, World Health Organization. Health statistics annual Geneva, Powles JW, Hage BH, Cosgrove M. Health related expenditure patterns in selected migrant groups: data from the Australian Household Expenditure Survey, Community Health Stud 199; : Powles JW, Hage BH, Ktenas DK. Who came? Who stayed behind? Selection and migration from the Greek island of Levkada. Departmental research report no. 88/1. Department of Social and Preventive Medicine, Monash University,, Australia, Powles JW, Ktenas D, Sutherland C, Hage B. Food habits in Southern-European migrants: a case study of migrants from the Greek island of Levkada. In: Truswell S, Wahlqvist M, eds. Food habits in Australia. Balwyn, Australia: Rene Gordon, 1988; McMichael AJ, McCall MG, Hartshorne JM, Woodings TL. Patterns of gastro-intestinal cancer in European migrants to Australia: role of dietary change. Intjf Cancer 198;: Australian Bureau of Statistics. National Health Survey, summary of results ABS cat. no , Australian Government Printing Services, Canberra, Australia, Australian Council on the Ageing and Department of Community Services. Older people at home, a report of a 1981 joint survey conducted in and Adelaide. Australian Government Printing Services, Canberra, Australia, Australian Institute of Multicultural Affairs. Community and institutional care for aged migrants in Australia: researchfindings. AIMA,, Australia, National Heart Foundation. Risk factor prevalence study no. 2, National Heart Foundation of Australia, Canberra, Australia, Trichopoulou A, Kouris-Blazos A, Vassilakou T, et al. The diet and survival of elderly Greeks: a link to the past. AmJClin Nutr 1995;61(6S): Kromhout D, Keys A, Aravanis C, et al. Food consumption patterns in the 196s in seven countries. Am J Clin Nutr 1989;49: Kafatos A, Kouroumalis I, Vlachonikolis I, Theodorou C, Labadarios D. Coronary heart disease risk-factor status for the Cretan urban population in the 198s. Amy Clin Nutr \99\;54:S9\ de Groot LCPGM, van Staveren WA, Hautvast JGAJ. Euronut-Seneca, nutrition and the elderly in Europe: a concerted action on nutrition and health in the European Community. Eur J Clin Nutr 1991;45(suppl 3): Trichopoulou A, Toupadaki N, Tzonou A, et al. The macronutrient composition of the Greek diet: estimates derived from six case control studies. Eur J Clin Nutr 1993;47: Trichopoulou A, Katsouyanni K, Gnardellis Ch. The traditional Greek diet. Eur J Clin Nutr 1993;47(suppl 1): Rutishauser I, Wahlqvist ML. Food intake patterns of Greek migrants to in relation to duration of stay. Proc Nutr Soc Aust 1983; Kosmidis G, Rutishauser I, Wahlqvist ML, McMichael A. Food intake patterns amongst Greek migrants in. Proc Nutr Soc Aust 198;5:5. 2. Department of Community Services and Health. National Dietary Survey of Adults, Canberra, Australian Government Publishing Service, Trichopoulou A, Efstathiadis PP. Changes in nutrition patterns and health indicators at the population level in Greece. AmJClin Nutr 1989;49: Wahlqvist ML, Davies L, Hsu-Hage BH-H, et al. Crosscultural study of food habits and health status in later life: description of elderly communities and lessons learnt. Asia Pacific J Clin Nutr and United Nations University Press, Wahlqvist ML, Kouris A. Trans-cultural aspects of nutrition in old age. Age Ageing 199:19(suppl):S Wahlqvist ML, Hsu-Hage B, Kouris-Blazos A, Lukito W. Food habits in later life: a cross-cultural study. In: Wahlqvist M, et al. eds. Nutrition in a sustainable environment. Proceedings of the XVth International Congress of Nutrition, United Kingdom: Smith- Gordon, 1994; Wahlqvist ML, Kouris A, Davies L, Scrimshaw N. Development of a Survey Instrument for the Assessment of Food Habits and Health in Later Life. In: Moyal M, ed. Dietetics in the 9s: role of the dietitian/nutritionist. Proceedings of the International Congress of Dietetics, Paris: J Libbey Eurotext, Kouris A, Wahlqvist ML, Trichopoulou A, Polychronopoulos E. Food habits and health of elderly in, Greece: application of Survey Instrument. In: New global era: global harmony through nutrition. Proceedings of the XlVth International Congress of Nutrition, United Kingdom: Smith-Gordon, Kouris-Blazos A. Elderly Greeks in, Greece and, Australia: food habits, health and lifestyle. PhD thesis. Monash University,, Australia, Kouris A, Wahlqvist ML, Trichopoulou A, Polychronopoulos E. Use of combined methodologies in assessing food beliefs and habits of elderly Greeks in Greece. Food Nutr Bull 1991;13: Australian Bureau of Statistics. The overseas-born in Victoria a statistical profile, ABS cat. no. 3.2, Australian Government Printing Services, Canberra, Australia, Lawton MP, Moss M, Fulcomer M, Kleban MH. A research and service oriented multilevel assessment instrument. J Gerontol 1982;37: Fillenbaum, GG. The wellbeing of the elderly: approaches to multidimensional assessment. Geneva: World Health Organization, Offset, Publication no. 84, Macrae FA, Lambert JR, Wahlqvist MC, Selbie L, Brouwer R, Maclennan R. The Australian Polyp Prevention Project: compliance to randomized dietary intervention. Proc Gastroenterol Soc Aust Trichopoulou A, Kampman E, Kolias E, Georga K. A photographic method to estimate food and nutrient intake in Greece-user's manual. Department of Nutrition and Biochemistry, Athens School of Public Health, Greece, Schofield WN, Schofield C, James WPT. Basal metabolic Downloaded from by guest on 17 November 218

13 HEALTH AND NUTRITIONAL STATUS OF ELDERLY GREEK MIGRANTS 189 rate-review, prediction, together with annotated bibliography source material. Hum Nutr Clin Nutr 1985;39C (suppl l):l Goldberg GR, Black AE, Jebb SA, Cole TJ, Murgatroyd PR, Prentice AM. Critical evaluation of energy intake data using fundamental principles of energy physiology: 1. Derivation of cut-off limits to identify under-recording. EurJCUn Nutr 1991;45: FAO/WHO/UNU. Energy and protein requirements. Report of a joint FAO/WHO/UNU Expert consultation. WHO Tech Rep Ser No. 74 Geneva: World Health Organization, National Research Council. Recommended dietary allowances. 1th edn. Washington DC: National Academy Press, Deurenberg P, van der Kooy K, Leenen R, Westsrate JA, Seidell JC. Sex and age specific prediction formulas for estimating body composition from bioelectrical impedance: a cross-validation study. Int J Obesity 1991 ;15: Gibson R. Principles of nutritional assessment. Oxford: Oxford University Press, Keys A, ed. Seven countries: a multivariate analysis of death and coronary heart diseases. Cambridge, Mass.: Harvard University Press, Trichopoulou A. Mediterranean diet, disease and nutrition guidelines. In: Benito E, Giacosa A, Hill MJ, eds. Public education on diet and cancer. (Proc 9th Annu Symp Eur Organ Coop Cancer Prev Stud) Kluwer Academic, 1992; Kafatos A, Kouroumalis I, Vlachonikolis I, Theodorou C, Labadarios D. Coronary heart disease risk-factor status for the Cretan urban population in the 198s. AmJ Clin Nutr 1991;54: Trichopoulos D, Skalkidis Y, Trichopoulou A, Touloumi O, Souliou L. The health of the Greek population: yesterday, today, tomorrow. Mater Med Greca 1989; Aravanis C, Ioannidis PJ. Nutritional factors and cardiovascular diseases in the Greek Islands Heart Study. In: Lovenberg W, Yamori Y, eds. Nutritional prevention of cardiovascular disease. New York: Academic Press, 1984; Trichopoulos D, Xirouchaki E, Tzonou A, Katsouyanni K, Zavitsanos X. Cancer in Greece: regional differences and secular trends. Mater Med Greca 1982; 1: Spiller GA. The Mediterranean diets in health and disease. Health Research and Studies Centre, Los Altos, California and Sphera Foundation, New York: Van Nostrand Reinhold, Wahlqvist ML, Kouris-Blazos A, Trichopoulou A, Polychronopoulos E. The wisdom of the Greek cuisine and way of life: comparison of the food and health beliefs of elderly Greeks in Greece and Australia. Age Nutr 1991;2: Harvard School of Public Health. International conference on the diets of the Mediterranean. Cambridge, Mass.: Oldways Preservation and Exchange Trust Publications, James WPT, Duthie G, Wahle K. The Mediterranean diet: protective or simply non-toxic? Eur J Clin Nutr 1989;43:(supp2): Baghurst KI, Crawford DA, Worsley A, Syrette JA, Record SJ, Baghurst PA. The Victorian Nutrition Survey: a profile of the energy, macronutrient and sodium intakes of the population. Community Health Stud 1988;22: Food and Agricultural Organization. Standardized food balance sheets. Rome: FAO, Wahlqvist ML, Kouris-Blazos A. Diet related disorders state of play. Food and Nutrition Policy, Department of Community Services and Health, Canberra, Australia, US National Research Council. Diet and health: implications for reducing chronic disease risk. Washington DC: National Academy Press, Slater R, Kritchevsky D. Human nutrition: a comprehensive treatise: cancer and nutrition. New York: Plenum Press, Wahlqvist ML, Kouris-Blazos A. Olive-oil nutrition and biological value. In: World Encylopedia of the Olive. International Olive Oil Council, Frankel EN, Kanner J, German JB, Parks E, Kinsella KE. Inhibition of oxidation of human low-density lipoprotein by phenolic substances in red wine. Lancet 1993;341: Byers TE. Diet and cancer: any progress in the interim? Cancer 1988;62: Keys A, Menotti A, Karvonen MJ, et al. The diet and 15 year death rate in seven countries. Am J Epidemiol 1986; 124: Horwarth CC. Dietary intake studies in elderly people. In: Bourne GH, ed Impact of nutrition on health and disease. World Rev Nutr Diet 1989;59:l Wahlqvist ML, Kouris-Blazos A, Lukito W, Hsu-Hage B. Water soluble vitamin intakes in the elderly: crosscultural findings in the IUNS study. In: Rosenberg IH, ed. Nutritional assessment of elderly populations: measure and function. New York: Raven Press, 1994; Roche AF. Anthropometric data. In: Wahlqvist ML, Davies L, Hsu-Hage BH-H, eds. Cross-cultural study of food habits and health status in later life: description of elderly communities and lessons learnt. Asia Pacific J Clin Nutr 1995 (in press). 62. Chandra RK. The relation between immunology, nutrition and disease in elderly people. Age Ageing 199; 19(suppl):S Stadtman ER. Protein oxidation and ageing. Science 1992;7: Authors' addresses A. Kouris-Blazos, M. L. Wahlqvist Department of Medicine, Monash Medical Centre, Monash University, Clayton Road, Clayton,, Victoria, Australia A. Trichopoulou, E. Polychronopoulos Department of Nutrition and Biochemistry, Athens School of Public Health, Greece D. Trichopoulos Department of Epidemiology, Harvard School of Public Health, USA Received in revised form 29 September 1995 Downloaded from by guest on 17 November 218