Dietary Intakes, Obesity and Health Behaviors in Native Hawaiians Residing in Southern California

Abstract

Objectives

Accessing dietary intakes, body mass index (BMI) and health behaviors in Native Hawaiians residing in Southern California.

Design

Cross-sectional, community based participatory research.

Participants

Native Hawaiian (N = 55); Mean age 59 (± 15).

Main Outcome Measures

Diet, body mass index (BMI), and diet/exercise health behaviors. Collected diet via 24-hr dietary recalls, health behaviors through questionnaires and BMI via measurement/self-report.

Analysis

Frequencies/means and multiple linear regression were used to assess diet, BMI, and health behaviors.

Results

Nearly 90% of the participants were either overweight or obese. Less than 20% met the vegetable, fruit, fiber, and whole grain recommendations. Most were a little or somewhat sure (relative to almost always sure) about their ability “to stick with an exercise program when attending a cultural gathering”, and “when visiting Hawai‘i”.

Conclusions and Implications

These results suggest that developing a culturally-based education program to reduce obesity and improve diet is critical for Native Hawaiians residing in Southern California.

Introduction

Native Hawaiians have the second highest overall incidence of cancer in Hawai‘i and second highest all-site cancer mortality rate in the United States.^1–3 Several lifestyle factors, including diet and obesity have been linked to cancer risk.^4–8 Studies on diet and obesity in Native Hawaiians show either higher energy intakes and/or increased body mass index (BMI).^9–13 The Multiethnic Cohort Study (MECS) reports that Native Hawaiians consume the lowest amount of legumes, but nutrient intakes from supplements are similar to the other ethnic groups.^18–19 These studies underscore the need for further investigation into strategies for improving nutrition and weight management for Native Hawaiians.

California has the largest Native Hawaiian and other Pacific Islander population (approximately 262,000), outside the state of Hawai‘i.¹⁴ However, culturally sensitive educational programs addressing cancer risk, obesity, and dietary intakes for Native Hawaiians have traditionally been conducted in Hawai‘i and not directed towards Native Hawaiians residing in Southern California.^15–18,20 While some cultural and behavioral issues, such as respect and caring for the family and one's health, are similar between the Hawaiians in Hawai‘i and in Southern California, Hawaiians in California have different structural needs in terms of access to food (eg, taro is more readily available in Hawai‘i), and knowledge related to diet.¹⁹ Our previous study of Native Hawaiians in California showed differences in psychosocial support by cardiometabolic status, however information on whether the community is meeting macro and micronutrient guidelines, and factors influencing culturally-specific health behaviors, which may impact future health programs for Native Hawaiians, have yet to be reported.²¹

Therefore, we conducted an exploratory community based participatory research (CBPR) study assessing diet, obesity, and health behaviors related to food and exercise for Southern California Native Hawaiians. Further, we wanted to assess whether culturally specific behaviors influence diet and exercise. To help reduce obesity and improve nutrition for Native Hawaiians, and subsequently develop culturally-appropriate health messages, information on dietary intakes and behaviors related to diet and exercise in Native Hawaiians should be made available and investigated.

Methods

Study Overview and Sample

Study population, recruitment, and study design have been described previously.²¹ Briefly, the present CBPR study was a cross-sectional study with a non-probability sample of Native Hawaiian adults ≥ 18 years of age residing in Southern California. Our key partners from ‘Āinahau O Kaleponi Civic Club and the Pacific Islander Health Partnership comprised the community advisory board (CAB) and were involved with all phases of the study from design, to recruitment, to results dissemination. Eligibility included ≥ 18 years of age, having some Native Hawaiian ancestry, and currently residing in Southern California. For the present study, community leaders and study staff, after obtaining written consent, collected two self-reported questionnaires (demographic and sociocultural), and in-person height and weight measurement at an initial assessment meeting. Following the collection of these first sets of data, three 24-hr dietary recalls via telephone were scheduled and collected. At subsequent assessment meetings self-reported health behavior data related to food and exercise were collected. Sixty-two recruited and consented individuals completed the demographic and socio-cultural questionnaires; 55 (88.7%) completed the 24-hr dietary recall and health behavior questionnaire, and constitute (N = 55) the present analysis. A full study protocol review was conducted and was approved by the Internal Review Board of California State University, Fullerton (HSR#: 09-0159).

Dietary Assessments

We assessed dietary and supplement intakes via 24-hr dietary recalls. Trained dietary assessors conducted three 24-hour dietary recalls stratified over weekdays and weekends during a 3-week period. Twenty-four hour dietary recall methodology uses multiple-pass, computer-assisted technology to collect all types of foods consumed and supplements used, over the previous 24-hour period for the participant.

The Nutrition Data System (NDS, University of Minnesota, Minneapolis, MN) used to conduct the 24-hour recalls is an interactive interface between the interviewer and participant, allowing for collection of detailed dietary/supplement intakes, including amounts, dosage, and brands. The nutrient database used by NDS to obtain the nutrients from food and supplements is derived from the USDA Nutrient Data Laboratory. NDS includes values for 144 nutrients, nutrient ratios, and food components and encompasses over 18,000 foods, including ethnic foods, and over 8,000 brand products. This is in direct contrast to other dietary assessment methodologies such as food frequency questionnaires or food checklists which are limited by a set-number and types of foods. Because NDS includes ingredient choices and preparation methods, more than 160,000 food variants are provided as options. Therefore, for different ethnic foods and recipes, as well as for collecting the most accurate and up-to-date dietary data, NDS is an appropriate tool for dietary assessments for Native Hawaiians and potentially for other Pacific Islander populations.

Health Behavior/Psychosocial Questionnaire

The health behavior instrument and psychometric evaluations of the scales used to assess behaviors related to diet and exercise have been validated and measured in diverse populations.^22–24 The health behavior questionnaire measures 15 scales: social support for diet, social support for exercise, self-monitoring diet, self-monitoring exercise, self-efficacy related to diet, self-efficacy related to exercise, outcome expectations for diet and for exercise, diet planning, preparing/buying foods, portion control, social interaction related to diet, social interaction related to exercise, and cognitive-behavioral strategies related to diet and exercise. The responses for all but one scale were on a 4 point Likert scale (1–4) ranging from “almost never to almost always” and “not at all sure to very sure”; the outcome expectations scale was on a 5 point Likert scale. Participants reported on their experiences over the past month.

Our CAB suggested revisions to the questionnaire that included culturally-specific questions such as: “how sure are you that you could stick to an exercise program in the following situation: when attending a cultural gathering (‘Ohana, Luau, New Year, Christmas)” and “when visiting Hawai‘i.” Additional questions are shown in Table 2.

Table 2.

Mean daily intakes of food groups, macro- and micronutrients and supplements for males and females (n = 55)

	Males (N=21)			Females (N=34)
	Mean (SD)	Range	% above guidelinesa,b	Mean(SD)	Range	% above guidelinesa,b
Food Intake
Energy kcal	1882 (710)	992–3220		1601 (615)	579–3010
% Energy from fat	33.7 (5.9)	22.3–44.2	24	34.8 (8.4)	18.2–48.3	22
Fiber, g	19.5 (6.3)	7.9–30.1	19	15.2 (7.3)	2.8–31.4	12
Vegetable, servings	2.9 (1.7)	0.7–6.6	14	2.3 (1.4)	0–5.8	3
Fruit, servings	1.9 (2.1)	0–8.3	14	1.5 (1.2)	0–5.6	3
Whole grain, servings	2.0 (2.0)	0–8.6	5	1.4 (1.3)	0–5.4	0
Refined grain, servings	4.0 (1.5)	0–7.3	66	3.5 (2.1)	0–9.0.	56
Total Folate, mg	452 (204)	240–995	52	393 (285)	104–1781	35
Natural Folate, mg	265 (125)	123–609	19	204 (89)	123–607	3
Synthetic Folate, mg	187 (166)	46–853	5	190 (236)	23–1373	9
Vitamin C, mg	110 (92)	28–322	48	83 (53)	23–205	44
Vitamin E, IU	10.3 (6.2)	3.4–28.0	19	13.4 (17.4)	1.8–100.9	18
Total carotenoids, mgc	10913 (6676)	1672–25118		8857 (6568)	487–25161
Calcium, mg	754 (459)	234–2433	10	714 (332)	129–1568	9
Iron, mg	15.5 (5.5)	8.9–32.2	100	13.6 (9.6)	4.4–61.2	79
Supplement Intake
Folate, mg	175 (246)	0–900	33	251 (309)	0–1000	44
Vitamin C, mg	168 (404)	0–1500	38	163 (330)	0–1060	26
Vitamin E, IU	94 (214)	0–833	43	65 (151)	0–667	47
Carotenoids, mgc	217 (598)	0–2550		1247 (6500)	0–38002
Calcium, mg	139 (272)	0–1120	0	367 (661)	0–2970	9
Iron, mg	2.4 (8.2)	0–36.3	10	3.8 (8.3)	0–36	18

^aGuidelines are based on AICR and USDA dietary recommendation: < 30% energy from fat (USDA), 25 – 30 g fiber (USDA), 5 servings of vegetable (AICR), 5 servings of fruit (AICR).⁶,³²

^bDRI guidelines for micronutrients have been previously reported.³³

^cPrecise dietary recommendations for carotenoids are unavailable.

Other Measures

We obtained sociodemographic data via a self-reported questionnaire. Height was measured in-person using a Seca 214 portable stadiometer (Hanover, MD), and weight was measured by an electronic step-up Ohaus ES200L bench scale (Pine Brook, NJ). Measurements were conducted using standardized guidelines previously reported.²⁵

Statistical Analysis

Descriptive analysis was conducted on gender, age (continuous), and education. Frequencies were also conducted on whether the participant was born, lived, or had family in Hawai‘i, and on supplement use. Height and weight data were used to calculate frequencies of normal (< 25 kg/m²), overweight (25–29.99 kg/m²), and obese (> 30 kg/m²) BMI. Mean intakes for food groups, macronutrients, and micronutrients from food and supplements were conducted. We calculated sample frequencies of participants who either met or exceeded dietary guidelines, for food groups and nutrients (based on Dietary Reference Intakes), established by the American Institute for Cancer Research (AICR) or the United States Department of Agricultural (USDA).^26–27 We conducted a t test and examined dietary intake differences between men and women. Cronbach's alpha coefficient was calculated for each of the 15 scales stratified by gender, including the culturally-specific questions. Means stratified by gender were calculated for each health behavior scale and individual culturally-specific questions.

Separate multivariate linear regression analysis were conducted to measure the association between each of the following independent dietary variables (independent), vegetable, fruit, fiber, and % energy from fat, and BMI with each health behavior scale (continuous dependent variable). Only those that completed all questions on all scales and questionnaires were included in the multivariate model (n = 37). Models were adjusted for age (continuous), gender, and education (no college education vs some college education). Significance for each test was set at P≤0.05. Analyses were conducted using SAS Version 9.1.

Results

Participation rates for completing all study components including questionnaires (health behavior and sociodemographic) and three 24-hr dietary recalls was 89%. Sociodemographic characteristics are presented in Table 1. The mean age for the current population was 59 (±15) and a majority of the participants were females (61.8%), 41.8% had some college education, and slightly over half (54.6%) used dietary supplements.

Table 1.

Anthropometrics and sociodemographic characteristic data for Native Hawaiians enrolled in the present study (N = 55)

Variable	Strata
Age (Mean ± SD)		59 ± 15 yrs
Body Mass Index, n (%)	Normal (19–24.9 kg/m2)	7 (12.7%)
	Overweight (25–29.9 kg/m2)	18 (32.7%)
	Obese (> 30 kg/m2)	30 (54.6%)
Education, n (%)	High School graduate	11 (20.0%)
	Some college	23 (41.8%)
	College Graduate	8 (14.6%)
	Advanced Degree	13 (23.6%)
Sex, n (%)	Female	34 (61.8%)
	Male	21(38.2%)
Supplement User, n (%)	Yes	30 (54.6%)
Born in Hawai‘i, n (%)	Yes	42 (76.4%)
	No	13 (23.6%)
Lived in Hawai‘i, n (%)	Yes	44 (80.0%)
	No	9 (16.4%)
	Unspecified	2 (3.6%)
Any family in Hawai‘i, n (%)	Yes	53 (96.4%)
	No	1 (1.8%)
	Unspecified	1(1.8%)

Table 2 presents the daily food groups, macro and micronutrient intakes from food and supplements, and dietary recommendation intakes stratified by gender. Data on frequency consumption related to the AICR/USDA dietary recommendations, for men, showed that <20% of the sample consumed above the recommendations for vegetable, fruit, fiber, whole grains, natural folate, vitamin E, and calcium. Similarly for women, <20% consumed above the recommendations for vegetable, fruit, fiber, whole grains, natural folate, synthetic folate, vitamin E, and calcium. However, between 30%–50% of both men and women were meeting guidelines for total folate, vitamin C, iron, and refined grains.

Individual questions specific to Native Hawaiian health behaviors related to food and exercise, presented in Table 3, show that most participants were either a little sure or somewhat sure about “sticking with an exercise program when attending a cultural gathering (‘Ohana, Luau, New Years)” and “when visiting Hawai‘i.” Mean scores for “how often did you choose leaner meat options and substitutes for lau lau and Kalua pig” were significantly higher for women (2.5 ± 1.1) compared with men (1.8 ± 0.9; P ≤ 0.05). Mean scores for “how satisfied are you with your current weight” was significantly lower for women (1.8 ± 1.0) compared with men (2.4 ± 1.0; P ≤ 0.05). No significant gender differences were found between mean scores for the thirteen scales on health behaviors related to diet and exercise (data not shown in tables). Briefly, for both men and women, the responses for most scales were at the intermediate level with mean scores for social support for exercise being the lowest (1.9 ± 0.7, 1.9 ± 0.6, respectively). The Cronhbach alpha coefficients ranged, for all 15 scales, from 0.71 to 0.96 for males, and from 0.79–0.95 for 13 out of 15 scales, for females, indicating good to excellent internal consistency. For females, two scales, portion control and social interaction related to diet, had a lower alpha of 0.64 and 0.69, respectively.

Table 3.

Means and standard deviations for individual culturally-specific health behavior questions related to diet and exercise for Native Hawaiians

Variable	Males		Females
	n	Mean(SD)	n	Mean(SD)
How sure are you that you can stick to an exercise program when attending a cultural gathering, (‘Ohana, Luau, New Year, Christmas)?a,d	19	2.8(1.0)	33	2.5(0.9)
How sure are you that you can stick to an exercise program when visiting Hawai‘i?a,d	19	2.8(1.0)	33	2.5(1.2)
How often did you find healthy Hawaiian foods readily available and accessible?b,e	19	2.0(0.9)	34	2.0(0.9)
How often did you choose leaner meats over those higher in fat, such as Spam, Portugese sausage, Vienna sausage?b,e	19	2.2(1.1)	34	2.4(1.0)
How often did you choose leaner meat options of substitutes for lau lau and Kalua pig?b,c,e	19	1.8(0.9)c	31	2.5(1.1)
How often did you prepare healthy foods with your family and friends?b,f	19	2.5(1.1)	33	2.5(0.9)
How satisfied are you with your current weight?b,c	19	2.4(1.0)c	33	1.8(1.0)

^aResponse options “not at all sure, a little sure, somewhat sure, very sure”; scale range 1 – 4

^bResponse options “almost never, sometimes, often, and almost always”; scale range 1 – 4

^cp < 0.05 between males and females

^dQuestions included in the self-efficacy related to diet scale

^eQuestions included in the planning related to diet scale

^fQuestion included in the preparation/buying scale

Multivariate analysis showed significant associations between dietary intakes and BMI with the diet/exercise health behavior scales (data not shown in tables). We found that fruit intake had a significant, positive association with healthful preparation and buying of foods (P ≤ 0.05; β = 0.872). Also, % energy from fat had a significant, positive association with higher social support for diet and higher outcome expectations related to diet (P ≤ 0.05; β = 0.11 and 0.38, respectively), but % energy from fat had a significant negative association with healthful preparation and buying of food (P ≤ 0.05; β = −0.2). BMI had a significant negative association with self-efficacy related to exercise, and outcome expectations related to diet and exercise (P ≤ 0.05; β = −0.16, −0.25 & −0.29, respectively).

Discussion

According to our findings, Native Hawaiians in Southern California are connected with Hawai‘i either through family or living and/or being born in Hawai‘i. Nearly 88% of the sample was either overweight or obese, and < 20% of the group met the vegetable, fruit, fiber, and whole grain recommendations. Finally, participants with lower self-efficacy related to diet and lower outcome expectations for diet and exercise had a higher BMI.

Studies have been conducted on dietary intakes and obesity in Native Hawaiians, primarily in Hawai‘i.^9–13,21,29 The MECS showed that Native Hawaiians in general had a high average fiber consumption (25.9 g), but they also consumed the most amount of energy (2,727 kcal) compared with Latinos (2,679 kcal), African Americans (2,278), Japanese-Americans (2,242) and Whites (2,340).²⁸ A recent study of 434 Native Hawaiians in North Kohala Hawai‘i also found that Native Hawaiians consumed the most amount of calories (2672.5 ± 1239.7).¹⁰ In contrast, energy intake was lower in our population which could be due to differences in dietary assessment methodology between studies. Further, our study shows that > 80% of the sample is not meeting AICR/USDA guidelines for vegetable, fruit, and fiber consumption, and similar to other reports, 87.3% were either overweight and/or obese.^10,13,29 These results suggest that Native Hawaiians in Southern California could benefit from further nutrition and weight management education/interventions related to cancer prevention.

Several health behaviors related to diet and exercise can be improved upon for Native Hawaiians. Social support for exercise should be encouraged. Because concepts such as ‘ohana (family) and working together are prevalent Native Hawaiian values,¹⁷ programs focusing on social support in a group setting may be well-received. Previous nutrition and health education programs in Hawai‘i using Native Hawaiian cultural values and integrating community and family social support have been effective.^17–18 Our findings suggest that increasing availability and accessibility of healthful Hawaiian foods, and particularly for men, providing alternatives for leaner substitutes for lau lau and Kalua pig, could initiate dietary behavior change. In addition, higher fruit and lower % energy from fat intake was associated with healthful preparation of foods. To improve Native Hawaiian dietary intakes, educational programs should focus on healthful preparation of foods, such as choosing leaner meats and cutting off visible fat. Moreover, we showed that lower BMI was associated with higher self-efficacy related to exercise and greater outcome expectations associated with diet and exercise. Thus, weight management programs for Native Hawaiians should focus on increasing self-efficacy related to adjusting meals by making them lower in fat, identifying low fat and low calorie foods, and developing programs on individual benefits of outcomes related to healthful eating and exercise such as feeling less depressed and losing weight. Attending to these issues could help motivate and encourage Southern California Native Hawaiians to make healthful dietary and weight management behavior modifications.

This study is one of the first studies to assess diet and health behaviors related to exercise/diet in Native Hawaiians residing in Southern California. However, limitations should be discussed. The small sample size may have limited detecting all potential associations in the multivariate model, observing differences between genders and generalizing the results. Despite the small sample size, we were able to detect significant associations between diet/BMI and health behaviors related to diet and exercise and gender differences. In addition, the study sample recruited from two primary community organizations in Southern California had a higher median age and education level, suggesting limitations related to representativeness of the results to the general Native Hawaiian population. Therefore, future studies should include a larger sample size and use population-based recruitment to ensure generalizability and representativeness. Nonetheless, the present study contributes to the sparse data on dietary intakes and health behaviors in Native Hawaiians residing in Southern California, and also demonstrates that joint partnership with the Native Hawaiian community, and incorporation of cultural values, can result in high participation rates and involvement of the community in research studies assessing diet and health behaviors related to nutrition and exercise.³⁰

Considering that Native Hawaiians are at high risk for cancer morbidity and mortality, developing a culturally-specific diet and weight management education program may reduce health disparities in this population. Importantly, the present study was initiated and mobilized by the Native Hawaiian community in Southern California which has significant implications for the effectiveness of future nutrition and physical activity programs for this community.

Conflict of Interest

The authors report no conflict of interest.