Abstract
ABSTRACT
Background
The Republic of Palau is facing a growing burden of non-communicable diseases (NCDs). Dietary diversity is a recognised indicator of diet quality and is linked to body mass index (BMI) and perceived health. However, its role in the rapidly changing food environment in Palau remains unclear. This study aimed to examine associations among dietary diversity, BMI and perceived health among adults and identify region-specific dietary patterns and food items linked to health outcomes.
Methods
A cross-sectional survey using structured door-to-door interviews was conducted between August and December 2023 with 263 Palauan adults aged 20 years and older across urban and rural areas. Data on dietary intake, anthropometry and self-rated health were collected. Dietary diversity was assessed using a modified 11-item Food Diversity Score Kyoto (FDSK-11), which categorised meat and fish into fresh and processed forms. Multiple and logistic regression analyses were performed, adjusting for sociodemographic covariates.
Results
FDSK-11 scores ranged from 3 to 11 (mean: 9.0), with no statistically significant difference between urban and rural areas. However, the consumption of root crops, fresh fish, processed meat and fruits was statistically higher in rural areas, while egg and dairy consumption was more common in urban areas. The obesity rate reached 58.8%, while 57.1% of participants rated their health as poor. The study found no statistically significant association between dietary diversity and BMI or perceived health. In rural areas, higher fruit intake was inversely associated with BMI, and root crop consumption was positively associated with better perceived health. In urban areas, higher intake of fresh fish, processed fish, dairy products and vegetables was linked to better perceived health.
Conclusion
These findings imply that promoting culturally relevant, high-quality local foods may be more effective than emphasising dietary diversity. Thus, public health strategies should prioritise the revitalisation of traditional food systems to address the increasing NCD burden in Palau.
Keywords: Dietary patterns, Malnutrition, Nutrition assessment, Metabolic syndrome
WHAT IS ALREADY KNOWN ON THIS TOPIC
Dietary diversity is widely recognised as a proxy indicator of diet quality and is typically associated with improved health outcomes such as lower body mass index (BMI) and better perceived health.
Pacific Island countries, including Palau, are experiencing a rapid increase in non-communicable diseases (NCDs), mainly driven by dietary shifts and food system transitions.
Previous studies in the Pacific have reported mixed results regarding the association between dietary diversity and NCD risk; less is known about these patterns in Palau.
WHAT THIS STUDY ADDS
This study is the first to examine associations among dietary diversity, BMI and perceived health in a nationally representative adult sample in Palau.
Although overall dietary diversity scores were high, they were not statistically significantly associated with BMI or perceived health. Instead, the consumption of specific foods (eg, fruits, root crops, vegetables and fish) exhibited significant associations with health outcomes.
Notable regional differences were observed in food consumption patterns despite the lack of difference in overall diversity scores, highlighting the influence of geography and cultural practices.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Public health strategies should focus on promoting dietary diversity and improving access to nutrient-dense, culturally relevant and minimally processed food.
Revitalising traditional food systems in Pacific Island countries such as Palau may be an effective approach to addressing the increasing burden of NCDs.
Future research should explore the qualitative aspects of dietary patterns and consider the incorporation of cultural and environmental factors into policies and intervention programmes for nutrition.
Introduction
Obesity is a leading cause of diet-related non-communicable diseases (NCDs) such as cardiovascular diseases, cancers and diabetes. In 2021, these conditions accounted for more than 70% of deaths globally, with a significant proportion occurring prematurely (under the age of 70 years), and more than 80% of NCD-related deaths reported in low- and middle-income countries. In response, the United Nations launched the Decade of Action on Nutrition (2016–2025) to address malnutrition, including obesity. Despite this effort, cases of obesity continue to increase, particularly in low- and middle-income countries, mainly due to inadequate access to affordable, nutritious food.1
The Republic of Palau, a small island nation with a population of approximately 16 733, is part of the Micronesian Islands in the Western Pacific and faces an increasing public health burden due to NCDs, which accounted for approximately 72% of deaths as of 2020.2 3 In 2020, a WHO survey reported that 42%, 50% and 32% of the population were obese, hypertensive and diabetic, respectively.4 Between 2012 and 2022, the prevalence of obesity increased from 26.1% to 38.2% among men and from 41.6% to 44.0% among women.4
These increases in obesity and diabetes are linked to food insecurity and dietary shifts. Palau features a complex colonial history, having been ruled by Spain, Germany and Japan for approximately 30 years after World War I. After World War II, it came under US trusteeship and gained independence in 1994. As a result, Palau’s food culture reflects influences from Micronesia, Japan, the USA and the Philippines. Since World War II, modernisation and urbanisation have transformed Palau’s food system. Historically, the Palauan diet was based on local root crops, fruits, seafood and vegetables. In recent decades, however, a sharp transition has been observed towards a westernised diet characterised by high-calorie, low-nutrient foods such as refined cereals (mainly rice) and imported processed foods.5 6 Frequently referred to as the ‘nutrition transition’, this transformation is a major driver of the increasing burden of obesity and NCDs in Palau. To date, Palau remains one of the least food-secure countries in the Pacific, with 85% of its total food consumption being imported.7 Nearly all grains are imported, whereas the local availability of fresh meat and vegetables is limited.6
Dietary diversity (the number of food items and food groups consumed over a given period) serves as an important indicator of diet quality. It reflects eating behaviours and is associated with perceived health and body mass index (BMI).8 High levels of dietary diversity have been linked to lower risks of obesity and micronutrient deficiencies.9 10 Studies conducted in Pacific Island nations, including Tonga, Samoa, Fiji, the Solomon Islands and Kiribati, have examined patterns of dietary diversity.11,17 In the Solomon Islands, for example, low dietary diversity was associated with micronutrient deficiencies and obesity.11 15
Scholars have demonstrated that perceived health (a subjective sense of well-being) is correlated with mortality risk.18 If dietary diversity is positively associated with perceived health and BMI, then public health interventions in Palau should focus on improving food quality. However, relevant research remains limited. Thus, the current study aims to evaluate the health status of adults in Palau, including obesity, perceived health and dietary habits, and to identify the association between dietary diversity and health indicators. The findings could contribute to improved dietary quality and inform effective public health strategies across the Pacific Islands.
Methods
Study design and participants
Palau has a population of approximately 16 733, with about 70% residing in Koror.19 This study conducted a cross-sectional survey from August to December 2023 with 263 participants from both urban and rural areas.
Urban participants were residents of Koror and Airai, while rural participants lived in Ngiwal, Ngchesar and Melekeok (eastern region); Ngardmau, Ngeremlengui and Aimeliik (western region); Ngaraard, Ngerchelong and Ngatpang (northern region); and Kayangel and Peleliu (small islands).
Data were collected via structured door-to-door interviews. Adults aged 18 years and older were invited to participate. Informed consent was obtained, and interviews were conducted in Palauan or English, depending on participant preference.
Only individuals who self-identified as Palauan were included in the study; foreign nationals, such as Filipinos and other ethnic groups, were excluded. All adult Palauan residents at home during the visit were invited. If no one was home, no data were collected and no follow-up visits were conducted. When multiple eligible individuals were present, all were invited to participate individually.
Anthropometric measurements
Height was self-reported, while weight was measured using a portable digital scale with a precision of 0.1 kg (BC-705N, TANITA, Japan). Participants were instructed to wear minimal clothing and to remove shoes and socks during measurement. The WHO standard BMI classification20 was applied to assess overweight and obesity. While alternative BMI cut-offs have been proposed for certain populations, we used standard thresholds to maintain consistency with previous research in Pacific Island settings.
Perceived health
Perceived health was assessed using the question, ‘How is your health in general?’, with the following response options: very bad, bad, fair, good or very good. This standardised item is recommended by the WHO and the European Union Survey on Income and Living Conditions.21 Responses of ‘very bad’ and ‘bad’ were categorised as ‘poor’, while ‘good’ and ‘very good’ were classified as ‘good’.
Dietary diversity
Dietary diversity was assessed using the 11-item Food Diversity Score Kyoto (FDSK-11) questionnaire, which evaluates nutritional intake without requiring detailed dietary recall or portion estimation.22 It includes 11 food categories recommended by the Japanese Ministry of Health, Labour and Welfare: grains, root crops, vegetables, meat, fish, dairy products, eggs, beans, seaweed, fruits and nuts. The tool has been validated in studies conducted in the Solomon Islands, China, Thailand and Japan.1122,24
Compared with other indicators of dietary diversity, such as the Dietary Diversity Score or the Food Variety Score, the FDSK-11 focuses on individual-level dietary variety over a longer recall period (6 months), rather than a 24-hour intake. This feature enables a more accurate depiction of habitual dietary patterns and makes it suitable for use in low-resource or community settings.
This study modified the FDSK to differentiate between fresh and processed forms of meat and fish due to the high consumption of processed food in Palau. In this context, ‘processed’ refers to food that underwent packaging or preservation procedures such as curing, salting, drying or canning. Processed meat included ham, sausages, corned beef, biltong, beef jerky and canned meat, while processed fish referred to canned fish, dried seafood and salted seafood. Traditional food was defined as minimally processed, locally sourced items such as root crops, fresh fish, vegetables and fruits.
Participants rated the frequency of consumption of each food category over the past 6 months: ‘hardly eat’ (0 days/week), ‘sometimes’ (1–2 days/week), ‘often’ (3–4 days/week) and ‘every day’ (6–7 days/week). ‘Hardly eat’ received a score of 0, while the other three options were scored as 1. For meat and fish, consumption of either the fresh or processed form (or both) more than once per week was scored as 1. The total FDSK score ranged from 0 to 11, with higher scores indicating greater dietary diversity.
Statistical analysis
Statistical analyses were conducted using the EZR package in R (V.4.3.2).25 Descriptive statistics were calculated for all variables. Student’s t-tests and χ2 tests were used to compare variables between urban and rural participants. One-way analysis of variance was used to compare continuous variables across regional categories.
Logistic and multiple linear regression analyses were performed to examine the associations between BMI or perceived health and dietary variables, including the total score for the FDSK-11 and intake of individual food categories. All models were adjusted for age, sex, marital status, level of education and occupation.
Missing data were noted for several variables, ranging from 1/263 (0.4%) for study area to 15/263 (5.7%) for fresh fish consumption. We used multiple imputation instead of treating missing values as zero, because the missing values were likely due to a combination of factors, including interviewer omissions, incomplete questioning, or the unsureness or unwillingness of participants to answer certain items. Multivariate imputation by chained equations was performed with 100 imputations, and results were pooled using Rubin’s rules.26 Statistical significance was set at p<0.05 (two-tailed).
Results
Sociodemographic characteristics of participants
Table 1 presents the sociodemographic characteristics of the participants. P values indicate comparisons between urban and rural areas. The mean age of the participants was 45.2 years (SD=15.5), with 93.1% having completed at least high school and 67.2% being employed.
Table 1. Sociodemographic characteristics of participants.
| Overall (n=263) | Rural (n=92) | Urban (n=171) | P value | |
|---|---|---|---|---|
| Mean age (SD) | 45.2 (15.5) | 46.7 (16.8) | 44.4 (14.7) | 0.251 |
| Sex, n (%) | ||||
| Men | 147 (56.5) | 50 (54.9) | 97 (57.4) | 0.793 |
| Women | 113 (43.5) | 41 (45.1) | 72 (42.6) | |
| Marital status, n (%) | ||||
| Married/partnership | 99 (37.8) | 34 (37.4) | 65 (38.0) | 0.699 |
| Single | 137 (52.3) | 46 (50.5) | 91 (53.2) | |
| Divorced | 26 (9.9) | 11 (12.1) | 15 (8.8) | |
| Education level, n (%) | ||||
| Non-formal education, primary school, secondary school | 18 (6.9) | 11 (12.2) | 7 (4.1) | 0.02 |
| High school or higher | 243 (93.1) | 79 (87.8) | 164 (95.9) | |
| Employment status, n (%) | ||||
| Unemployed | 86 (32.8) | 37 (40.7) | 49 (28.7) | 0.054 |
| Employed | 176 (67.2) | 54 (59.3) | 122 (71.3) |
P values are based on Student’s t-test for FDSK-11 score and χ2 tests for categorical variables.
FDSK-11, 11-item Food Diversity Score Kyoto.
No statistically significant differences were observed between urban and rural participants in terms of age, sex, marital status or employment status. However, a statistically significant difference was found in the level of education (p=0.020), with a greater proportion of rural participants having only non-formal, primary or secondary education compared with urban participants (12.2% vs 4.1%).
Although the main comparison focused on urban and rural groups, additional regional differences were noted. Participants from the Small Islands were statistically significantly older (mean age=62.5 years), had lower levels of formal education (41.2% had not completed high school) and exhibited a markedly higher rate of unemployment (88.2%) compared with other regions. In contrast, Airai had the youngest participants, the highest educational attainment and the lowest unemployment rate.
Dietary diversity scores
Table 2 presents the FDSK-11 scores, which ranged from 3 to 11, with a mean score of 9.0 (SD=1.9). No statistically significant differences were observed between urban and rural areas. Table 2 also depicts the daily consumption rates of individual food items, with several statistically significant differences noted between the two groups. For example, the daily consumption of root crops, fresh fish, processed meat and fruits was statistically significantly higher in rural areas. In contrast, the consumption of egg and dairy products tended to be more common in urban areas, although these differences were not statistically significant.
Table 2. Scores in Food Diversity Score Kyoto and daily consumption pattern of each food by regional group.
| Overall (n=263) | Rural (n=92) | Urban (n=171) | P value | |
|---|---|---|---|---|
| FDSK-11 score, mean (SD) | 9.0 (1.9) | 8.79 (2.30) | 9.12 (1.75) | 0.202 |
| Daily consumption of each food, n (%) | ||||
| Grain | 185 (70.6) | 67 (73.6) | 118 (69.0) | 0.098 |
| Root crops | 71 (27.2) | 46 (50.5) | 25 (14.7) | <0.001 |
| Fresh meat | 116 (46.4) | 45 (50.6) | 71 (44.1) | 0.34 |
| Processed meat | 64 (24.7) | 31 (34.1) | 33 (19.6) | 0.022 |
| Fresh fish | 108 (43.5) | 51 (57.3) | 57 (35.8) | 0.004 |
| Processed fish | 48 (18.8) | 19 (21.1) | 29 (17.6) | 0.662 |
| Egg | 50 (20.0) | 14 (15.7) | 36 (22.4) | 0.091 |
| Dairy | 51 (19.7) | 12 (13.3) | 39 (23.1) | 0.081 |
| Vegetable | 90 (35.0) | 31 (34.8) | 59 (35.1) | 0.188 |
| Seaweed | 8 (3.1) | 2 (2.2) | 6 (3.6) | 0.085 |
| Beans | 10 (3.9) | 5 (5.6) | 5 (3.0) | 0.161 |
| Nuts | 22 (8.6) | 8 (9.0) | 14 (8.4) | 0.293 |
| Fruits | 58 (22.6) | 30 (33.7) | 28 (16.7) | 0.019 |
P values are based on Student’s t-test for FDSK-11 score and χ2 tests for categorical variables.
FDSK-11, 11-item Food Diversity Score Kyoto.
Further regional variation was evident within rural areas. For instance, the northern region exhibited notably high consumption rates of processed meat (51.7%) and fresh meat (74.1%), while the remote islands demonstrated a low intake of processed foods (processed meat: 15.9%; processed fish: 5.9%) but a higher consumption of fresh fish (68.8%).
Health status
Table 3 indicates that no statistically significant differences exist between urban and rural areas in terms of perceived health or BMI-related indicators. In summary, 57.1% of the participants rated their health as ‘bad’, with similar proportions observed in rural (59.1%) and urban (56.1%) areas. In terms of BMI, more than half of the participants (58.8%) were classified as obese, and only 13.6% fell under the normal weight category. The mean BMI was 32.2, with urban participants displaying slightly higher values than rural participants, although the difference was not statistically significant.
Table 3. General health status of participants by region.
| Overall (n=263) | Rural (n=92) | Urban (n=171) | P value | |
|---|---|---|---|---|
| Perceived health, n (%) | ||||
| Good | 111 (42.9) | 36 (40.9) | 75 (43.9) | 0.692 |
| Bad | 148 (57.1) | 52 (59.1) | 96 (56.1) | |
| BMI composition, n (%) | ||||
| Normal | 35 (13.6) | 14 (15.4) | 21 (12.7) | 0.469 |
| Overweight | 71 (27.6) | 21 (23.1) | 50 (30.1) | |
| Obesity | 151 (58.8) | 56 (61.5) | 95 (57.2) | |
| BMI, mean (SD) | 32.2 (8.6) | 31.2 (6.4) | 32.7 (9.6) | 0.181 |
P values are based on Student’s t-test for FDSK-11 score and χ2 tests for categorical variables.
BMI, body mass index; FDSK-11, 11-item Food Diversity Score Kyoto.
A further breakdown by subregion also revealed no statistically significant differences but highlighted certain geographical patterns. For instance, the proportion of participants who reported ‘good’ health was highest in Airai (55.3%), whereas it was notably low in East (26.7%), where 73.3% rated their health as ‘bad’. Regarding obesity, the highest rates were observed in East (71.4%) and Airai (69.4%), whereas the lowest rate was found in the Small Island region (52.9%). Similarly, mean BMI was highest in Airai (35.5) and Koror (32.5), and lowest in Small Islands (29.9) and North (30.2).
Table 4 reveals that among participants who reported good health, 65.1%, 22.9% and 11.9% were classified as obese, overweight and normal weight, respectively. In contrast, among those who reported poor health, 53.5%, 31.9% and 14.6% were obese, overweight and normal weight, respectively. However, the study found no statistically significant difference in BMI distribution between participants with good and poor perceived health.
Table 4. Cross-tabulation of perceived health and BMI categories (overall).
| BMI category | Good health n (%) |
Bad health n (%) |
P value |
|---|---|---|---|
| Normal | 13 (11.9) | 21 (14.6) | 0.170 |
| Overweight | 25 (22.9) | 46 (31.9) | |
| Obese | 71 (65.1) | 77 (53.5) |
P values were calculated using the χ2 test.
BMI, body mass index.
Correlations among FDSK-11 score, food consumption and health indicators
The FDSK-11 score was not statistically significantly associated with BMI in either the rural or urban study areas (table 5). Similarly, the consumption of most individual food items showed no significant association with BMI. However, a statistically significant inverse association was observed between fruit consumption and BMI in rural areas (β=−3.64, SD=1.37, p=0.009), indicating that greater fruit intake may be linked to lower BMI among rural participants.
Table 5. Multiple regression analysis of BMI with each food item.
| Food item | Urban β (SD) |
P value | Rural β (SD) |
P value |
|---|---|---|---|---|
| Grain | 1.92 (2.62) | 0.464 | 0.18 (2.20) | 0.932 |
| Root crops | −1.00 (1.61) | 0.534 | −0.55 (1.77) | 0.755 |
| Fresh meat | 0.95 (1.34) | 0.477 | 0.44 (1.81) | 0.806 |
| Processed meat | 0.98 (1.59) | 0.540 | 0.64 (1.58) | 0.689 |
| Fresh fish | −1.15 (1.16) | 0.320 | 1.52 (2.02) | 0.453 |
| Processed fish | −1.81 (1.20) | 0.134 | −0.84 (1.48) | 0.572 |
| Egg | −0.18 (1.56) | 0.906 | −0.98 (1.48) | 0.506 |
| Dairy | 0.79 (1.52) | 0.605 | −2.09 (1.47) | 0.159 |
| Vegetable | −1.90 (1.71) | 0.267 | −1.06 (1.49) | 0.481 |
| Seaweed | 3.78 (2.20) | 0.080 | −0.54 (1.76) | 0.761 |
| Beans | −1.77 (1.89) | 0.350 | −1.90 (1.70) | 0.266 |
| Nuts | −1.02 (1.79) | 0.568 | −1.48 (1.56) | 0.345 |
| Fruits | −2.43 (1.54) | 0.117 | −3.64 (1.37)* | 0.009* |
| FDSK-11 score | −0.47 (0.44) | 0.281 | 0.11 (0.31) | 0.705 |
Age, sex, level of education, occupation and marital status were adjusted.
*P<0.05 indicates statistically significant results.
BMI, body mass index; FDSK-11, 11-item Food Diversity Score Kyoto.
Regarding perceived health, the FDSK-11 score also showed no statistically significant association in either the urban (p=0.09) or rural (p=0.689) populations (table 6). However, several specific food items were statistically significantly associated with perceived health.
Table 6. Adjusted logistic regression analysis of the association between perceived health and each food item.
| Food item | Urban OR (95% CI) |
P value | Rural OR (95% CI) |
P value |
|---|---|---|---|---|
| Grain | 0.61 (0.19 to 1.97) | 0.407 | 2.82 (0.64 to 12.4) | 0.169 |
| Root crops | 1.29 (0.64 to 2.56) | 0.471 | 5.81 (1.56 to 21.6)* | 0.008* |
| Fresh meat | 1.43 (0.63 to 3.23) | 0.388 | 0.40 (0.08 to 2.04) | 0.272 |
| Processed meat | 1.53 (0.77 to 3.02) | 0.224 | 2.22 (0.74 to 6.54) | 0.153 |
| Fresh fish | 2.12 (1.01 to 4.46)* | 0.046* | 0.40 (0.08 to 2.04) | 0.272 |
| Processed fish | 2.18 (1.11 to 4.31)* | 0.025* | 1.66 (0.60 to 4.58) | 0.324 |
| Egg | 1.36 (0.69 to 2.65) | 0.373 | 0.64 (0.22 to 1.81) | 0.403 |
| Dairy | 1.97 (1.02 to 3.83)* | 0.044* | 0.40 (0.13 to 1.23) | 0.110 |
| Vegetable | 2.35 (1.09 to 5.03)* | 0.028* | 0.56 (0.20 to 1.57) | 0.270 |
| Seaweed | 0.88 (0.34 to 2.25) | 0.790 | 1.03 (0.30 to 3.51) | 0.958 |
| Beans | 1.20 (0.53 to 2.71) | 0.656 | 0.82 (0.24 to 2.74) | 0.752 |
| Nuts | 0.88 (0.41 to 1.89) | 0.757 | 0.83 (0.28 to 2.46) | 0.749 |
| Fruits | 1.60 (0.81 to 3.14) | 0.175 | 1.32 (0.50 to 3.48) | 0.575 |
| FDSK-11 score | 1.18 (0.96 to 1.43) | 0.090 | 1.05 (0.84 to 1.30) | 0.689 |
Age, sex, level of education, occupation and marital status were adjusted.
*P<0.05 indicates statistically significant results.
FDSK-11, 11-item Food Diversity Score Kyoto.
In rural areas, only the consumption of root crops was statistically significantly associated with better perceived health (adjusted OR=5.81, 95% CI 1.56 to 21.6, p=0.008). In contrast, in urban areas, better perceived health was statistically significantly associated with higher consumption of fresh fish (OR=2.12, 95% CI 1.01 to 4.46, p=0.046), processed fish (OR=2.18, 95% CI 1.11 to 4.31, p=0.025), dairy products (OR=1.97, 95% CI 1.02 to 3.83, p=0.044) and vegetables (OR=2.35, 95% CI 1.09 to 5.03, p=0.028).
No other food items exhibited statistically significant associations with either BMI or perceived health in either region.
These results suggest that, although overall dietary diversity (as measured by the FDSK-11 score) was not statistically significantly associated with BMI or perceived health, certain food items did show significant associations with these outcomes. Moreover, the pattern of associations varied between rural and urban areas.
Discussion
This study investigated the relationships among dietary diversity, food consumption and health indicators in Palauan adults. The findings revealed the complex influence of cultural background, lifestyle and social structure on dietary behaviour and health.
The study found no statistically significant differences between urban and rural areas in dietary diversity score (FDSK-11), BMI or perceived health. This finding suggests that living environment and food access may be relatively homogeneous across regions due to national infrastructure development and widespread reliance on imported food.27 The lack of association between dietary diversity and BMI or perceived health further implies that dietary content and health status may be relatively uniform across the country.
However, the study noted a number of regional differences in the consumption of specific foods. In rural areas, the intake of root crops, fresh fish, processed meat and fruits was higher, while the consumption of eggs and dairy products was more common in urban areas (although statistically non-significant). These differences likely reflect the influence of rural cultural practices such as home gardening, fishing and traditional food sharing.
Notably, a higher rate of fruit consumption in rural areas was statistically significantly associated with lower BMI, which is consistent with prior studies reporting that fruit intake enhances satiety, lowers total energy intake and contributes to decreased adiposity due to high fibre and water content and low energy density.28 In addition, fruits may impose a lower metabolic burden compared with energy-dense processed foods, thus potentially supporting effective weight regulation.29 Additionally, the consumption of root crops and fish (including processed fish) was positively associated with perceived health. Although processed food is typically considered a health risk, the measure of perceived health in this study included physical, mental and social aspects, which indicates that the convenience and familiarity of these foods may have positively contributed. Furthermore, in rural contexts with limited resources, processed food may offer reliable access to calories, thereby supporting a sense of security and well-being despite their lower nutritional value.
The combined overweight and obesity rate in our study reached 86.4%, which differed from a report by the Ministry of Health and Human Services in 2023 (overweight: 33.2%, obese: 42.4%). This discrepancy may be due to the composition of the study population, which consisted mainly of ethnic Palauans, among whom the prevalence of obesity is higher. A previous study reported overweight or obesity among 83.4% of Palauan adults compared with 58.2% of Filipinos and 53.8% of other ethnicities.3 Thus, the findings are internally valid despite the small sample size.
Despite the high prevalence of obesity, the findings revealed no clear association between BMI and perceived health status. Many participants who were classified as obese reported feeling in good health. This discrepancy may reflect cultural perceptions of body weight in Pacific Island settings. Ulijaszek30 argued that in a number of Pacific communities, increased body weight has historically been interpreted as a sign of prosperity and well-being rather than a health concern. In these contexts, individuals may not associate excess body weight with poor health unless accompanied by physical symptoms or functional impairment. This cultural framing may partially explain why BMI was not aligned with self-rated health in the current study population.
The mean FDSK-11 score reached 9.0 (range: 3–11) without statistically significant differences between urban and rural areas. Compared with other countries—Solomon Islands (7.5), Thailand (8.4) and China (7.9)1222,24—Palau displayed a relatively high score, which likely reflects its integration into global food systems and broad access to imports. However, despite high scores, the intake rates of vegetables and fruits remained low: only 35.0% of the participants consumed vegetables daily, while a mere 22.6% consumed fruits daily. These figures reflect previous studies that reported the majority of adults in Palau do not meet recommended intake levels.3
Cultural and structural factors likely underlie these low intake rates. Vegetables have not traditionally held a central role in Palauan cuisine; for example, the term yasay for vegetables is derived from Japanese. Many vegetables that are currently available are imported or newly introduced varieties,6 and barriers include limited local production, high prices and unfamiliarity. Palau remains one of the least food-secure countries in the Pacific, with 85% of its food derived from imports.7 Nearly all grains are imported, while fresh meat and vegetables are scarce.6 While rural residents maintain greater access to traditional foods, this has not necessarily led to improved diet quality or health. In fact, in a number of food categories, rural residents consumed more processed foods than did urban residents, which potentially reflects the urbanisation of rural lifestyles.
These findings indicate that dietary diversity scores alone may be inadequate for capturing dietary quality or health outcomes. Even with high diversity, health benefits may remain limited if nutrient-rich food is lacking. Especially given that the consumption of traditional, minimally processed foods exhibited stronger associations with lower BMI and better perceived health, public health interventions should emphasise the variety and the quality and cultural relevance of food.
Traditional Palauan diets were based on root crops (eg, taro, breadfruit, cassava and sweet potatoes), leafy greens (eg, taro and sweet potato leaves), marine products (eg, fish and shellfish) and fruits (eg, bananas, papayas, mangoes and wax apples). These foods are generally lower in fat and sodium and richer in complex carbohydrates, fibre, protein and micronutrients.31,33 Traditional cooking methods, such as boiling, steaming and roasting, are also healthier than frying,34 which poses clear nutritional advantages compared with modern Western-style diets.
Promoting traditional food practices could provide a culturally appropriate and nutritionally effective strategy for preventing NCDs in Palau. Ongoing initiatives by the Cooperative Research and Extension Department of Palau Community College and the Ministry of Agriculture, Fisheries, and Environment aim to strengthen local food systems.35 In particular, expanding the production of traditional starch crops may decrease dependence on imports, enhance food security and promote healthy eating habits. Future public health efforts should move beyond increasing diversity alone to supporting culturally and nutritionally sustainable food choices across the population.
Limitations
This study has its limitations. First, height and dietary intake were self-reported, which may introduce recall or social desirability bias. The participants may have over-reported or under-reported certain behaviours, which could affect the accuracy of BMI calculations and dietary patterns. Second, while the FDSK-11 is a cost-effective and simple tool for assessing dietary diversity, it does not account for serving size, portion control or nutrient density; therefore, it may only partially capture the quality or adequacy of diets. Third, the sample size was relatively small and primarily composed of Palauan individuals, which may limit the generalisability of the findings to the broader, multiethnic population of Palau. Fourth, although adults present in households were invited to participate, instances occurred in which only one household member responded. This scenario may have introduced gender bias if, for example, women were more likely to be at home during the survey hours and, therefore, were over-represented among the respondents.
Despite these limitations, the study provides valuable insights into the relationship between patterns of food consumption and health perceptions, particularly in the context of a small island developing state. Future studies with larger and more diverse samples, as well as objective dietary assessments, are required to validate and expand on the current findings.
Conclusion
This study found no statistically significant differences between urban and rural populations in terms of dietary diversity score (FDSK-11), BMI or perceived health. Additionally, dietary diversity was not significantly associated with BMI or perceived health. These non-findings suggest that dietary patterns and health outcomes in Palau may be relatively uniform across regions, likely due to similar lifestyles and equal access to imported foods across the nation.
However, the study observed statistically significant differences in the consumption of specific food items between urban and rural participants. The rural participants reported higher daily intake rates of root crops, fresh fish, fruits and processed meat, while the consumption of eggs and dairy products was more common in urban areas, although not statistically significant. Furthermore, the study noted an inverse association between fruit consumption and BMI in rural areas, and the intake of root crops, fish and even processed fish was positively associated with perceived health across both settings.
These findings indicate that the quality and cultural relevance of specific foods may be more important than dietary diversity in influencing eating behaviours and health perceptions. Therefore, future health promotion strategies in Palau should emphasise the consumption of traditional starches and minimally processed protein sources that are culturally appropriate and nutritionally beneficial, rather than adhering strictly to general recommendations for dietary diversity.
Acknowledgements
We would like to express our gratitude to the participants and the staff of Palau Community College for their invaluable support.
The funder had no role in the study design, data collection, analysis, interpretation, writing of the manuscript, or decision to submit the article for publication.
Footnotes
Funding: This study was funded by the KAKENHI Grant-in-Aid for Scientific Research (project numbers 20H00045 and 24H00123 (TF, Kyoto University)) from the Japan Society for the Promotion of Science, Japan.
Data availability free text: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Patient consent for publication: Not applicable.
Ethics approval: This study involves human participants and was approved by the Ethics Committee of Kyoto University (project number R4047). This study adhered to the Declaration of Helsinki. Participants gave informed consent to participate in the study before taking part.
Provenance and peer review: Not commissioned; externally peer reviewed.
Data availability statement
Data are available upon reasonable request.
References
- 1.United Nations General Assembly United nations decade of action on nutrition (2016–2025). Resolution A/70/259. 2016. [30-Jul-2025]. https://www.un.org/nutrition/sites/www.un.org.nutrition/files/general/pdf/conceptnote_mtr_nutrition_decade_0.pdf Available. Accessed.
- 2.Ministry of Health and Human Services. Republic of Palau . Koror: Ministry of Health and Human Services; 2015. [30-Jul-2025]. Non-communicable disease prevention and control strategic plan of action 2015–2020: healthy communities in a healthy palau.https://extranet.who.int/countryplanningcycles/planning-cycle-files/palau-non-communicable-disease-prevention-and-control-strategic-plan-action Available. accessed. [Google Scholar]
- 3.Ministry of Health and Human Services. Republic of Palau . Available from. Koror: Ministry of Health and Human Services; 2023. [30-Jul-2025]. PALAU adult hybrid/community health assessment survey 2023.https://ghdx.healthdata.org/record/palau-adult-hybridcommunity-health-assessment-survey-2023 Available. accessed. [Google Scholar]
- 4.World Health Organization Prevalence of obesity among adults, BMI ≥ 30 (age-standardized estimate) (%) 2024. [30-Jul-2025]. https://www.who.int/data/gho/data/indicators/indicator-details/GHO/prevalence-of-obesity-among-adults-bmi-=-30-(age-standardized-estimate)-(-) Available. Accessed.
- 5.Shimizu H, Hirakawa Y, Chiang C, et al. Perceptions and behaviors related to noncommunicable diseases in Palau: a qualitative study. Nagoya J Med Sci. 2021;83:287–98. doi: 10.18999/nagjms.83.2.287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Nakamura S, Iida A, Nakatani J, et al. Global land use of diets in a small island community: a case study of Palau in the Pacific. Environ Res Lett. 2021;16:065016. doi: 10.1088/1748-9326/ac0212. [DOI] [Google Scholar]
- 7.JIRCAS . Agriculture in Palau. Japan: Japan International Research Center for Agricultural Sciences; 2016. The situation of current agricultural production in palau; pp. 52–9.https://www.jircas.go.jp/en/publication/agriculture_in_palau Available. [Google Scholar]
- 8.Appannah G, Emi NA, Gan WY, et al. The Relationships between a Dietary Pattern Linked to Cardiometabolic Risk Factors and Life Satisfaction in Early Adolescence. Int J Environ Res Public Health. 2020;17:5489. doi: 10.3390/ijerph17155489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Seyoum Keflie T, Samuel A, Lambert C, et al. Dietary Patterns and Risk of Micronutrient Deficiencies: their Implication for Nutritional Intervention in Ethiopia. J Nutrit Health Food Sci. 2018;6:1–16. doi: 10.15226/jnhfs.2018.001120. [DOI] [Google Scholar]
- 10.Tao C, Zhao Q, Glauben T, et al. Does Dietary Diversity Reduce the Risk of Obesity? Empirical Evidence from Rural School Children in China. Int J Environ Res Public Health. 2020;17:8122. doi: 10.3390/ijerph17218122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Tsuchiya C, Furusawa T, Tagini S, et al. Socioeconomic and Behavioral Factors Associated with Obesity Across Sex and Age in Honiara, Solomon Islands. Hawaii J Health Soc Welf . 2021;80:24–32. https://hawaiijournalhealth.org/past_issues/80.02.htm Available. [PMC free article] [PubMed] [Google Scholar]
- 12.Tsuchiya C, Tagini S, Cafa D, et al. Socio-environmental and behavioral risk factors associated with obesity in the capital (Honiara), the Solomon Islands; case-control study. Obesity Medicine . 2017;7:34–42. doi: 10.1016/j.obmed.2017.07.001. [DOI] [Google Scholar]
- 13.DiBello JR, McGarvey ST, Kraft P, et al. Dietary Patterns Are Associated with Metabolic Syndrome in Adult Samoans. J Nutr. 2009;139:1933–43. doi: 10.3945/jn.109.107888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Eme PE, Burlingame B, Douwes J, et al. Quantitative estimates of dietary intake in households of South Tarawa, Kiribati. Asia Pac J Clin Nutr. 2019;28:131–8. doi: 10.6133/apjcn.201903_28(1).0018. [DOI] [PubMed] [Google Scholar]
- 15.Horsey B, Swanepoel L, Underhill S, et al. Dietary Diversity of an Adult Solomon Islands Population. Nutrients. 2019;11:1622. doi: 10.3390/nu11071622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Konishi S, Watanabe C, Umezaki M, et al. Energy and nutrient intake of Tongan adults estimated by 24-hour recall: the importance of local food items. Ecol Food Nutr. 2011;50:337–50. doi: 10.1080/03670244.2011.586315. [DOI] [PubMed] [Google Scholar]
- 17.O’Meara L, Williams SL, Hickes D, et al. Predictors of Dietary Diversity of Indigenous Food-Producing Households in Rural Fiji. Nutrients. 2019;11:1629. doi: 10.3390/nu11071629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kaplan GA, Camacho T. Perceived health and mortality: a nine-year follow-up of the human population laboratory cohort. Am J Epidemiol. 1983;117:292–304. doi: 10.1093/oxfordjournals.aje.a113541. [DOI] [PubMed] [Google Scholar]
- 19.Government of Palau 2020 census of population and housing report. [30-Jul-2025]. https://www.palaugov.pw/wp-content/uploads/2022/09/2020-Census-of-Population-and-Housing.pdf Available. Accessed.
- 20.World Health Organization Obesity and overweight. 2025. [30-Jul-2025]. https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight Available. Accessed.
- 21.Eurostat EU-SILC – European Union statistics on income and living conditions. 2024. [30-Jul-2025]. https://ec.europa.eu/eurostat/web/microdata/european-union-statistics-on-income-and-living-conditions Available. Accessed. [DOI] [PubMed]
- 22.Kimura Y, Ogawa H, Yoshihara A, et al. Evaluation of chewing ability and its relationship with activities of daily living, depression, cognitive status and food intake in the community-dwelling elderly. Geriatr Gerontol Int. 2013;13:718–25. doi: 10.1111/ggi.12006. [DOI] [PubMed] [Google Scholar]
- 23.Kimura Y, Okumiya K, Sakamoto R, et al. Comprehensive geriatric assessment of elderly highlanders in Qinghai, China IV: Comparison of food diversity and its relation to health of Han and Tibetan elderly. Geriatrics Gerontology Int. 2009;9:359–65. doi: 10.1111/j.1447-0594.2009.00543.x. [DOI] [PubMed] [Google Scholar]
- 24.Tsuchiya C, Amnatsatsue K, Sirikulchayanonta C, et al. Lifestyle-related factors for obesity among communitydwelling adults in Bangkok, Thailand. J Glob J Health. 2017;32:9–16. https://jglobal.jst.go.jp/en/detail?JGLOBAL_ID=201702248634646491 Available. [Google Scholar]
- 25.Kanda Y. Investigation of the freely available easy-to-use software “EZR” for medical statistics. Bone Marrow Transplant. 2013;48:452–8. doi: 10.1038/bmt.2012.244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rubin DB, Schenker N. Multiple imputation in health-care databases: an overview and some applications. Stat Med. 1991;10:585–98. doi: 10.1002/sim.4780100410. [DOI] [PubMed] [Google Scholar]
- 27.JICA Data collection survey for the bridge sector in republic of Palau: final report. [30-Jul-2025]. https://openjicareport.jica.go.jp/pdf/12367355.pdf Available. Accessed.
- 28.Lin B-H, Morrison RM. Higher fruit consumption linked with lower body mass index. Food Rev. 2002;25:28–32. doi: 10.22004/ag.econ.234628. [DOI] [Google Scholar]
- 29.Slavin JL, Lloyd B. Health benefits of fruits and vegetables. Adv Nutr. 2012;3:506–16. doi: 10.3945/an.112.002154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ulijaszek SJ. Obesity and body image in Pacific populations. Asia Pac J Clin Nutr. 2007;16:220–4. [Google Scholar]
- 31.Kaufer L, Englberger L, Cue R, et al. Evaluation of a “traditional food for health” intervention in Pohnpei, Federated States of Micronesia. Pac Health Dialog. 2010;16:61–73. https://pubmed.ncbi.nlm.nih.gov/20968237 Available. [PubMed] [Google Scholar]
- 32.Dela Cruz R, Wolfe E, Yonemori KM, et al. Consumption of Traditional Fruits and Vegetables among Children in the US-Affiliated Pacific Region. Curr Dev Nutr. 2022;6:nzac101. doi: 10.1093/cdn/nzac101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Shintani TT, Hughes CK. Traditional diets of the Pacific and coronary heart disease. J Cardiovasc Risk. 1994;1:16–20. doi: 10.1097/00043798-199406000-00004. [DOI] [PubMed] [Google Scholar]
- 34.McGregor A. The Palau PACC food security project: a benefit cost analysis. 2012. [30-Jul-2025]. https://www.adaptation-undp.org/sites/default/files/resources/palau_pacc_cba_final_report.pdf Available. Accessed.
- 35.Solutions for a Food Secure PALAU Secretariat of the Pacific Regional Environment Programme (SPREP) 2022. [30-Jul-2025]. https://www.sprep.org/news/solutions-for-a-food-secure-palau Available. Accessed.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data are available upon reasonable request.