Table 4 Characteristics and main findings of studies assessing fish consumption^a

Hedges et al. [54]

20 non-pregnant non nursing females in Verata

Cross sectional study :

a) Food records - food and beverages consumed over two x weekly periods, 3 weeks apart

Energy derived from protein, fat and carbohydrate (CHO), intake of protein and sources of protein intake

Mean energy from protein/carbohydrate/fat over the 2 weeks was 13 %/66 %/21 % respectively.

Mean intake of energy was 9080 KJ/day.

Mean intake of protein was 70.6 g. Primary sources of protein were cereals (3.7 % of protein energy) followed by fish (3.4 %), meat (1.7 %) and shell fish (1.5 %). Higher consumption of marine foods was associated with lower cereal intake (−0.54).

Kuster et al. [45]

Ono-i-Lau Island

30 senior heads of households in 1982

59 households in 2002

Cross sectional

Mean weekly household income and fish yield per capital.

Daily fish intake per capita (g) and contribution of marine sources to protein intake.

Total annual landings of finfish decreased by 27 % from 1982 to 2002

No significant change in yield of finfish per capita of population (96.9 kg per capita/year in 1982 to 93.7 kg capita/year in 2002)

Seafood remained the main source of protein between 1982 and 2002.

Consumption of canned fish increased from 9 g/man/day to 19 g/man/day

Turner et al. [42]

Lau Province, 25 experienced local fishers

53 senior heads of households

Cross sectional

Semi structured face-to-face interviews

Time spent fishing, importance of fishing for income generation, patterns of fish consumption and awareness of ecological change within the local qoliqoli (fishing ground)

No significant change in overall time spent fishing in the past 6–10 years.

Income-generating activities had increased in importance over previous10 years relative to fishing activities Consumption of fresh fish was significantly lower compared to estimates of past consumption (Z = −3.774, p < 0.001).

Greatest decline in fish consumption was associated with highest mean household income.

Of the 80–100 % of households that engaged in fishing, only 7 % of households ranked fishing as the primary household occupation.

Guillemot et al. [44]

146 local fishers on the North west coast

Cross sectional study:

Face-to-face questionnaire

Average number of fishing trips per year, average catch per fishing trip (kg) and annual reef fish consumption per capita

Total catches of reef fish estimated at 169 t/year.

7720 fishing trips made/year.

Mean catch per fishing trip per boat was 23 kg.

Annual reef fish consumption estimated at 18 kg/person/year

Labrosse et al. [50]

646 individuals aged over 7 years, in the Northern Province

Cross sectional study:

Face-to-face 13-item questionnaire

Weekly and annual fish intake, quantity of fish per meal (g), subsistence versus purchased fish intake

Only 1.3 % of the participants reported never eating fish. 85 % of participants ate fish 1 or more times a week, with 45 % of these eating fish 2–3 times/week and 11 % consuming fish 1–2/day.

Average weekly consumption of fish was 4.8 meals ± 0.7.

Average quantity of fish consumed at a meal was 233 g ± 16 g.

92 % percent of annual fish consumption was classified as subsistence with only 8 % purchased.

Leopold et al. [48]

38 households on Ouvea Island

Cross sectional study:

Face-to-face household questionnaire

Annual fish consumption per household and per consumption unit (CU)

Annual fish consumption per household was a mean 289 kg.

Mean fish consumption per CU was 63 kg ± 9.7 kg (SD).

Subsistence production was the main source of supply of fish for two thirds of households, followed by gifts and purchases from the market.

Aswani & Furusawa [52]

Five villages in the Roviana Lagoon,

106 adult males and females 574 males and females aged ≥15 years

437 males and females ≥18 years

(231 in 2001 and 206 in 2005)

Cross sectional

Face-to-face semi-structured interviews, 24 h dietary recall

Anthropometric assessment and 1 h dietary recall questionnaire

Contribution of marine resources to protein intake

Energy (MJ) and protein (g) intake

64 – 100 % of all participants identified fish as the main source of protein consumed by their household

Males and females consumed a sufficient amount of protein according to the FAO/WHO/UNU reference values 1985.

Between 2001 and 2005 mean energy intake in both males (P < 0.01) and females (P < 0.0001) increased. There were no statistically significant changes in BMI for either males or females between 2001 and 2005.

Mertz et al. [41]

46 heads of households in Tikopia

Cross sectional

Face-to-face household questionnaire

Fish and meat intake;

Fish consumed daily by 72 % of households.

Main source of dietary protein was fish

Imported canned fish and meat were eaten when available, but was rarely available.

van der Heijden [57]

609 individuals residing in high, middle and low altitude villages in the Ramu catchment and 204 fishers from the Yonki Reservoir (high altitude)

Cross sectional study:

Structured face-to-face interviews

Contribution of fish to intake of animal based foods

Fish and other aquatic animals caught in nearby waters contributed 7.7 % of food from animal sources for high and middle altitude participants.

Fish was the most important source of animal protein for low altitude respondents (41 % of all food from animal sources) and the second most important source for respondents at Yonki Reservoir (25 % of all food from animal sources).

Canned fish and lamb flaps were the most commonly purchased animal foods

Vengiau et al. [34]

70 adult Naasioi migrants residing in Port Moresby aged 18–65 years

Cross sectional study:

Face-to-face 39-item FFQ (foods were classified as ‘traditional foods’ or ‘store foods’)

Assessment of household socio-economic status determined by fortnightly income, educational level, house type and number of household possessions

Association between socio-economic status and food intake type

“Socio-economic index” was positively correlated with greater consumption of “traditional foods” e.g., bananas, tubers, fruits and vegetables compared to “store foods” e.g., rice, noodles, canned fish and soft drink (P = 0.04)

Vengiau et al. [35]

70 adult Naasioi migrants residing in Port Moresby aged 18–65 years

Cross sectional study:

Face-to-face 39-item FFQ to establish key dietary patterns

Anthropometric assessment including height (cm), weight (kg), and blood pressure (BP)

Identification of key dietary patterns

Association between dietary pattern and BMI (kg/m²) and blood pressure (mm Hg)

Two dietary patterns emerged. ‘Traditional diet’ - which was correlated with consumption of bananas, tubers, other sweet potato, pawpaw and leafy greens and negatively correlated with consumption of rice, canned fish and soft drinks. The second dietary pattern was correlated with consumption of store bought foods e.g., rice, noodles, canned fish and soft drinks

Median BMI was 27 in females and 29 in males. 38 % of females and 23 % of males classified as obese (BMI ≥ 30).

Systolic BP was > than 140 mm Hg in 8 % of females and 19 % of males

Diastolic BP was > than 90 mm Hg in 10 % of females and 13 % of males.

No association was found between cardiovascular risk factors (BMI and BP) and dietary patterns

Dancause et al. [46]

425 children and 559 adults on three islands varying in economic development (Ambae, Aneityum and Efate)

Cross sectional study:

Face-to-face behavioural questionnaire

24 h dietary recall

Behavourial changes within and among islands associated with economic development

Hypertension or CVD among family members was reported by 19.7 % of participants in Ambae, by 35.3 % in Aneityum and 45.1 % in Efate.

Overweight and obesity amongst family members was reported by 24.8 % of participants in Ambae, 48.0 % in Aneityum and 39.6 % in Efate.

Fresh fish intake ranged from approximately 10 % in Ambae to 50 % in Aneityum. Meat and/or fish intake was highest in Efate, followed by Aneityum then Ambae (P < 0.001). Tinned meat was consumed only monthly or yearly by the majority of participants in Ambae and Aneityum, compared to consumption daily or weekly for 80 % of participants in Efate.

Children in Efate reported taking more processed foods to school compared to children from Aneityum and Ambae who reported taking more local foods to school.

Dancause et al.[59]

534 males and females aged ≥18 years on three islands: Ambae, Aneityum and Efate

Cross sectional study:

Face-to-face behavioural questionnaire

24 h dietary recall questionnaire

Anthropometric assessment including height, weight (kg),

Association between fish and meat intake and anthropometric measures

For males and females means of anthropometric measures were lowest in Ambae (rural), intermediate in Aneityum (rural with tourism) and highest in Efate (suburban).

Risk factors for obesity included consuming canned fish (OR 2.91, 95 % CI; P = 0.020) and consuming multiple fish/meat dishes (OR 3.24, 95 % CI; P = 0.010)

Li et al. [39]

153 school children aged 8–10 years from four locations on the island of Tanna

Cross sectional study:

Face-to-face dietary survey

Spot urine sample

Thyroid volume

Intake of fish and type of fish consumed

Median urinary iodine excretion (UIE), iodine sufficiency and association between fish consumption iodine sufficiency

Association between thyroid volume and UIE

39 % of children reported eating fish on at least a weekly basis and 44 % on a monthly basis. Canned fish consumption was reported by 70 % of participants.

Eating fish at least monthly was associated with urinary iodine sufficiency (UIE > 100 ug/L, P = 0.011; P = 0.045 after adjusting for location, age and sex).

Children were moderately iodine deficient and had much larger thyroid glands compared to international reference values.A statistical significant inverse correlation between thyroid volume and UIE for boys (r = −0.444, P = 0.001; n 77) and girls (r = −0.319, P = 0.005; n 76)

Polynesia

Clero et al. [36]

229 cases :203 women and 26 men diagnosed with thyroid cancer from 1979 to 2004

371 controls: matched to cases by sex and age

Case–control study:

66-item FFQ

Energy intake (kcal), iodine nutrition status and risk of thyroid cancer

Cases consumed a mean 71 g/day of fish and shellfish compared to 83 g/day in controls.

Iodine nutrition deficiency (<150 ug/day) was observed in 60 % of cases and controls. 30 % had optimal iodine nutrition (150–299 ug/day).

Risk of thyroid cancer decreased significantly with increasing consumption of fish (P = 0.008) and shellfish (P = 0.002). Subjects with a severe or moderate iodine nutrition deficiency had a 2.6 times increased risk of thyroid cancer (95 % CI: 1.12, 5.93) compared to subjects with optimal iodine nutrition status.

Clero et al.[33]

229 cases :203 women and 26 men diagnosed with thyroid cancer from 1979 to 2004

371 controls: matched to cases by sex and age

Case–control study:

66-item FFQ

Association between dietary pattern and thyroid cancer risk

2 major dietary patterns were identified: Western and traditional Polynesian.

The traditional Polynesian pattern was inversely associated with risk of thyroid cancer. After adjustment for total energy intake, a 42 % reduced risk of thyroid cancer was found for the highest vs. the lowest tertile: (OR −0.58; 95 % CI: 0.35, 0.95; P = 0.02). The decreased risk was attenuated after multivariate adjustment (p = 0.2)

Dewailly et al.[38]

214 pregnant women who gave birth between October 2005 and February 2006 and their neonates

Cross sectional study:

Face-to-face survey

Blood sample taken from the umbilical cord of neonates

Monthly fish intake (meals/month) including type of fish

Iodine (I), selenium (Se) and omega 3 polyunsaturated fatty acid (n-3 PUFA) concentrations

In pregnant women, mean fish consumption was 33 meals/month of which 21.3 and 11.5 meals/months were from reef and pelagic fish respectively. Tuna (75 %) was the most common pelagic fish species consumed.

Mean umbilical iodine cord blood concentrations varied between different archipelagos. Highest mean concentration of iodine was 2.60 umol/L and the lowest was 0.46 umol/L. Mean cord blood selenium concentration was 2.0 umol/L. Mean concentration of n-3 PUFAs in red cell membrane phospholipids was 3.52 %. Eicosapentanoic acid (EPA) and docosahexanoic acid (DHA) represented 87 % of all n-3 PUFAs

Konishi et al. [53]

19 females (aged 40–59 years) and their spouses (15) from Kolovai village

Cross sectional study:

Face-to-face 24 h dietary recall administered over 7 consecutive days during two different seasons

Anthropometric assessment of weight and height

Protein intake (g) and contribution of local fish and imported meats to nutrient intake

Weight (kg), height (cm) and BMI (kg/m²)

Daily protein intake was 112 g and 97 g for men and women respectively. 97 % of participants were considered to have adequate protein intake.

Fish contributed 20 % toward total protein intake compared to the sum of the two major imported meats mutton and chicken (23 % of total protein intake). Mutton (23 %) and chicken (10 %) were the highest sources of fat in the diet.

Mean BMI was 32.3 kg/m² ± 4.4 kg/m² and 36.3 kg/m² ± 5.4 kg/m²for men and women respectively

Kronen & Bender [43]

Lofanga Island

Individual adults (>15 years), households, key informants, 41 local fisherman

Mixed methods

Households and individuals: structured questionnaires.

Key informant interviews

Fishermen: Semi-structured interviews

Fishing practices including contribution of fish to livelihood and weekly fish consumption through

Fisheries, agricultural production and handicrafts all contributed to income generation. Fisheries ranked higher than agriculture and handicrafts.

Fisheries filled three main objectives: to secure subsistence, fulfil social obligations and contribute to the cash economy. Fish^a was consumed by entire community. 93 % of village also consume other seafood and 78 % consume canned fish.

Smith et al. [49]

443 school students aged 11–16 years from Tongatapu, Vava’u and Hapa’ai.

Cross sectional study:

Self-administered health behaviour survey

Anthropometric assessment including height (cms) and weight (kgs)

Intake of canned fish, mutton and beef

Prevalence of overweight and obesity

35 % of participants reported consuming canned fish once or more a day compared to 57 % who reported consuming canned mutton or beef once or more a day

The prevalence of overweight and obesity was 36 % amongst boys and 54 % amongst girls using international cut-off points for BMI.

Craig et al. [47]

594 fishermen and 20 males aged 43–72 years from Ofu, Olosega and Sili villages in American Samoa

Cross sectional

Observation of fishing activities and semi-structured interviews with fishermen

Free-form interviews of 20 males

Annual fish consumption

Per capita fish catch was 71 kg.

63 kg of catch was consumed per capita per annum.

Remaining catch was either sold locally (60 % of fishers sold catch at least occasionally) or sent to family members on Tutuila Island.

DiBello et al. [58]

723 American Samoans and 785 Samoans aged ≥18 years

Cross sectional study:

Anthropometric assessment (weight (kg), height (cm) and waist circumference). Fasting blood samples and blood pressure were measured.

FFQ (42-item American Samoa and 55-item Samoa)

Prevalence of metabolic syndrome

Association between dietary pattern and metabolic syndrome

The prevalence of metabolic syndrome was 49.4 % and 30.6 % in American Samoan and Samoan samples respectively

‘Neo traditional’ and ‘modern’ dietary patterns derived. The ‘neo-traditional’ pattern was characterised by high intake of crab and lobster, fish, coconut cream dishes, papaya soup, coconut milk, papaya.Significant increasing prevalence of metabolic syndrome across increasing quintiles of the ‘modern’ dietary pattern in Samoa (P = 0.05).

The ‘neo-traditional’ dietary pattern was associated with an increase in HDL cholesterol in America Samoa (P = 0.02) and decreased waist circumference in both communities (P = 0.03).

Corsi et al. [37]

293 females aged 15–64 years located in Ponhpei

Cross sectional study:

27-item, 7-day FFQ

Knowledge, attitudes and practices questionnaire

Fish and meat consumption

Cash expenditure on food; factors affecting food intake

79 % of participants reported frequent consumption of local fish/seafood. Local fish/seafood was consumed twice as frequently (4.8 days/week) compared to imported fish/seafood (2.4 days/week). Imported meats such as turkey tail were consumed more frequently 1.9 days/week) than local meats (1.3 days/week).

8 % of participants reported their household relied on farming and fishing for their primary income 6 % relied on fishing alone.

Household food expenditure for 77 % of participants was half or more of their monthly income

52 % of participants purchased local food for half or more than half of a month.

Consuming imported food was regarded as a sign of wealth and status by participants

Englberger et al.[55]

Kosrae

267 children aged 24–59 months and their caretakers 65 children aged 24–59 months and their caretakers

Cross sectional study:

34-item 7-day FFQ

Three non-consecutive 24 h dietary recall questionnaires

Intake of total Vitamin A, retinol and protein (g). Sources of vitamin A

Main dietary components included imported products of rice, flour, chicken, other meats and tinned fish and local products of breadfruit, banana, taro, fish and other seafood.

The most frequently consumed protein food was imported frozen chicken, followed by local fish.

Mean intake of protein (54 g ± 10 g), well above the estimated requirements.

Mean daily intake of total vitamin A for all children was less than half of the requirements estimated by WHO and the FAO^#. Animal sources provided 52-53 % of vitamin A intake for both children and caretakers.

Pobocik et al. [51]

211 males and 189 females

Cross sectional study:

Multiple-pass telephone 24 h dietary recall

Self reported weight (kg) and height (m)

Fish and other meat intake

BMI (kg/m²)

Most commonly reported meats consumed were chicken (reported 159 times), beef (148), fish (141), eggs (75), sausage/bacon/hot dogs (65), pork/ham (60) and canned meat (42).

Canned and processed meats accounted for 21 % of all reported meat intake. When fish was consumed 47 % was fresh and the remaining was canned or dried.

Mean BMI was 25.7 kg/m² ± 5.8 kg/m² with significant differences in BMI by ethnicity (P < 0.05)

Phongsavan et al. [40]

4885 school children aged 13–15 years from Republic of Vanuatu, Kingdom of Tonga and Federated States of Micronesia

Cross sectional study:

Self-administered Health Behaviour and Lifestyle of Pacific Youth (HBLPY) survey.

Canned fish and mutton intake

In Vanuatu, canned fish was consumed on a daily basis by 40 % of students and canned mutton 17 % of students.

In Tonga 32 % of students consumed canned fish on a daily basis and 52 % of students consumed canned mutton daily.

In Pohnpei students were asked about consumption of fresh fish - 42 % reporting consuming fresh fish on a daily basis and 46 % consumed canned mutton daily