Supplemental Material can be found at: http://jn.nutrition.org/content/suppl/2014/05/11/jn.113.18837 5.DCSupplemental.html

The Journal of Nutrition Nutritional Epidemiology

Healthy Dietary Habits Score as an Indicator of Diet Quality in New Zealand Adolescents1–3 Jyh Eiin Wong,4,6 Paula M. L. Skidmore,4 Sheila M. Williams,5 and Winsome R. Parnell4*

Abstract Adoption of optimal dietary habits during adolescence is associated with better health outcomes later in life. However, the associations between a pattern of healthy dietary habits encapsulated in an index and sociodemographic and nutrient intake have not been examined among adolescents. This study aimed to develop a behavior-based diet index and examine its validity in relation to sociodemographic factors, nutrient intakes, and biomarkers in a representative sample of New Zealand (NZ) adolescents aged 15–18 y (n = 694). A 17-item Healthy Dietary Habits Score for Adolescents (HDHS-A) was developed based on dietary habits information from the 2008/2009 NZ Adult Nutrition Survey. Post hoc trend analyses were used to identify the associations between HDHS-A score and nutrient intakes estimated by single 24-h diet recalls and selected nutritional biomarkers. Being female, not of Maori or Pacific ethnicity, and living in the least-deprived socioeconomic quintile were associated with a higher HDHS-A score (all P < 0.001). HDHS-A tertile was associated positively with intake of protein, dietary fiber, polyunsaturated fatty acid, and lactose and negatively with sucrose. Associations in the expected directions were also found with most micronutrients (P < 0.05), urinary sodium (P < 0.001), whole blood (P < 0.05), serum (P < 0.01), and RBC folate (P < 0.05) concentrations. This suggests that the HDHS-A is a valid indicator of diet quality among NZ adolescents. J. Nutr. 144: 937–942, 2014.

Introduction Healthful dietary habits are associated with better nutrient intake and higher diet quality, which in turn leads to positive health outcomes (1–3). Because dietary habits are likely to track into adulthood (4), adoption of optimal dietary habits during adolescence may have a protective effect against chronic diseases later in life (5). Food habits questionnaires have been used to collect qualitative information on food behaviors of adolescents, including food types, food preparation, cooking practices, snacking patterns, and intake frequency of certain food groups, which can then be compared with dietary guidelines (6–9). Compared with quantitative reporting of food intake, usual dietary habits are perceived to be more easily and accurately documented (9). Because dietary habits are interrelated and tend to cluster (2,10), there is a potential synergetic effect of multiple dietary habits, so it is important to examine them as a group rather than studying them in isolation (10). However, dietary habits information has seldom been examined comprehensively by means of diet index scores. In addition, the association of a constellation of desirable dietary habits, as assessed by a diet index, with sociodemo-

graphic, nutrient intake, and biomarker data are less clear because few studies allow investigation of these relations. To date, 12 studies of children and adolescents used an index embedding information on dietary habits. They mostly examine the association of the index with food or nutrient intakes (11– 16), body composition measures (11–19), and 1 with nutritional biomarkers (11). They were conducted only in convenience samples of children and adolescents aged 9–24 y. None of them examined the relation of a dietary habits index with nutrient intakes and nutritional biomarker levels in a nationally representative sample of adolescents. The aims of this study are 2-fold: 1) to develop Healthy Dietary Habits Scores for Adolescents (HDHS-A)7 based on dietary habits information; and 2) to examine the validity of the HDHS-A based on its internal reliability and associations with sociodemographic factors, nutrient intakes, and nutritional biomarkers in adolescents aged 15–18 y who participated in the 2008/2009 New Zealand (NZ) Adult Nutrition Survey.

Methods 1

Supported by grants from the New Zealand Ministry of Health. Author disclosures: J. E. Wong, P. M. L. Skidmore, S. M. Williams, and W. R. Parnell, no conflicts of interest. 3 Supplemental Tables 1 and 2 are available from the ‘‘Online Supporting Material’’ link in the online posting of the article and from the same link in the online table of contents at http://jn.nutrition.org. * To whom correspondence should be addressed. E-mail: winsome.parnell@ otago.ac.nz. 2

Study design and sample. The study design included a secondary analysis of the 2008/2009 NZ Adult Nutrition Survey, a cross-sectional population-based survey in a representative sample of New Zealanders 7 Abbreviations used: DHQ, dietary habits questionnaire; HDHS-A, Healthy Dietary Habits Score for Adolescents; NZ, New Zealand; NZDep2006, 2006 New Zealand Deprivation Index; NZEO, New Zealand European and others.

ã 2014 American Society for Nutrition. Manuscript received November 14, 2013. Initial review completed January 5, 2014. Revision accepted March 14, 2014. First published online April 17, 2014; doi:10.3945/jn.113.188375.

937

Downloaded from jn.nutrition.org at UNIV OF CALIFORNIA SAN FRANCISCO KALMANOVITZ LIBRARY on January 17, 2015

Departments of 4Human Nutrition and 5Preventive and Social Medicine, University of Otago, Dunedin, New Zealand; and 6School of Healthcare Sciences, Faculty of Healthy Sciences, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia

aged $15 y. The methodology for this national nutrition survey was granted ethical approval from the NZ Health and Disability Multi-Region Ethics Committee and was described in detail previously (20). Briefly, a multistage, stratified, probability-proportional-to-size sampling was used, and 4721 home interviews were conducted from October 2008 to October 2009. Data from those aged 15–18 y were included in the present study.

Development of the HDHS-A. The HDHS-A index was developed using information from the DHQ after consultation with 2 expert nutritionists and a dietitian. The index items were selected on the basis that they were relevant to and captured the key nutrients that are important in determining the diet quality of NZ adolescents, as guided by existing scientific evidence and the NZ Food and Nutrition Guidelines for Healthy Children and Young People (21). Four index prototypes were initially created. Using 21 questions from the DHQ, the final prototype included 17 items grouped into 5 clusters: 1) fat from meat, poultry, and fish; 2) other fats; 3) fruit, vegetables, and bread; 4) sugar sources; and 5) meal habits. These clusters were named after main nutrients or domains of diet reflected by the healthy dietary habits. A response that aligned with a more positive dietary habit was assigned a higher score using a 5-point scoring system ranging from 0 to 4. The total HDHS-A was a summation of scores from the 17 items and ranged from 0 to 68. A greater total score represents a dietary pattern reflective of healthier dietary habits. The 17 items derived from the DHQ and their respective scoring criteria are presented in Supplemental Table 1. Sociodemographic information. Age was derived from the date of birth and interview start date. Participants were asked to identify $1 ethnic groups to which they belonged, and participants were classified hierarchically into 3 ethnic groups: 1) Maori; 2) Pacific; and 3) NZ European and others (NZEO) (25). Socioeconomic status was estimated using an area-based scale of deprivation, namely the 2006 NZ Deprivation Index (NZDep2006) (26). This proxy measures deprivation based on 8 dimensions of deprivation for each neighborhood (i.e., a mesh block containing ;87 people) in NZ, including income and benefit receipt, home ownership, support for sole-parent families, employment status, qualifications, 938

Wong et al.

Anthropometric measurements. Heights and weights were measured in the participantÕs home by trained interviewers using a standardized protocol (20). Standing height was measured to the closest 0.1 cm using a portable stadiometer (Seca 214; Seca), and body weight was measured in minimum clothing to the nearest 0.1 kg using an electronic weighing scale (model HD-351; Tanita). BMI was calculated by dividing weight in kilograms by height in meters squared. Weight status was defined using the International Obesity Taskforce age- and sex-specific BMI cutoffs for children and adolescents (27,28). Biochemical assessment. Within 2 wk after the home interviews, nonfasting blood and spot midstream urine samples were obtained from consenting participants at affiliated laboratories in local areas across NZ. Blood and urine biomarkers of dietary status included in this study were as follows: 1) whole-blood folate; 2) serum folate; 3) RBC folate; 4) urinary iodine; 5) urinary sodium; and 6) urinary potassium. Blood was collected from a forearm vein into 2 4-mL vacutainers (EDTA1 and EDTA2) containing EDTA and 1 10-mL vacutainer with no additive. All samples were stored at 4°C until analyzed. Complete blood count was determined from EDTA1 at local laboratories. After transportation to the Department of Human Nutrition, University of Otago, whole-blood folate was analyzed from EDTA2 using a microbiologic assay on 96-well microtiter plates with chloramphenicol-resistant Lactobacillus casei as the test microorganism (29). Using the same method, serum folate was analyzed using serum separated from vacutainers without additive. RBC folate was calculated by subtracting serum folate from whole-blood folate and correcting for hematocrit (30). RBC concentration of folate was shown previously to be a good indicator of long-term folate status (31). Spot urine samples were analyzed at the Canterbury Health Laboratories for urinary sodium, potassium, and iodine. Urinary sodium and potassium were analyzed using an ion-selective electrode analyzer with integrated chip technology (Abbott Architect C8000 biochemical analyzer), whereas urinary iodine was determined by the SandellKolthoff colorimetric method (32). Statistical analysis. Participants aged 15–18 y who completed a single 24-h diet recall and at least 75% of the DHQ (i.e., 19 of 25 questions) were considered eligible for the analysis (n = 695). Data for eligible participants with 1–3 missing responses (n = 103), entered either as ‘‘did not answer’’ or ‘‘donÕt know,’’ were estimated from other information that the participant supplied as long as at least 50% of questions related to the items within a cluster were completed. Mean values of nonmissing items in a cluster were estimated and substituted for the missing values (1 participant was excluded because >50% of the data were missing in 1 of the clusters). Because the data were part of a nationwide survey, the Stata survey procedures, which account for the complex structure of the sample and the sampling weights, were used to analyze the data. These provide the correct SE for the estimates. A natural logarithm was applied to transform nonnormally distributed data. Continuous variables were presented as means or geometric means if log-transformed and SEs or 95% CIs. Categorical data were presented as relative frequencies and percentage. Survey postestimation tests (lincom command in Stata) were used to compare continuous variables (e.g., age, BMI Z-score, HDHS-A score) between the sexes. All analyses were performed using the statistical software package Stata (version 11.2; StataCorp), with statistical significance set at P < 0.05. Evaluation of the HDHS-A. Content validity of the HDHS-A was established through an expert review to ensure that all 17 items of the index were important dietary habits contributing to the diet quality of NZ adolescents. The index was also tested for internal reliability using correlations and CronbachÕs a coefficient. To examine construct validity, 3 hypotheses were generated with regard to the associations between HDHS-A scores and sociodemographic factors, 24-h nutrient intakes, and nutritional biomarkers. It was hypothesized that higher HDHS-A

Downloaded from jn.nutrition.org at UNIV OF CALIFORNIA SAN FRANCISCO KALMANOVITZ LIBRARY on January 17, 2015

Dietary intake. Dietary intake was assessed using a dietary habits questionnaire (DHQ) and an interview-administered 24-h diet recall. The DHQ containing 25 questions was used to collect information on key eating habits associated with diet quality and nutritional status (21,22). This questionnaire comprised frequency questions on the past 4-wk consumption of breakfast, breads, meats (red meat, chicken, processed meat products, fish, or shellfish), fast foods or takeaways, french fries, sugary beverages, and confectionery. The other questions focused on daily intake of servings of fruit and vegetables and dietary practices pertaining to the use of fat and salt, e.g., removing excess fat from meat and adding salt to food on the plate before consumption and their usual choices of types of bread, milk, fat spread, and cooking fat. The participantsÕ responses were directly entered into computer-assisted personal interview software (Abbey Research Software, Life in New Zealand Health and Activity Research Unit, University of Otago). The DHQ was cognitively tested for suitability in assessing usual eating habits (20), and each component was used previously in other large nationally representative studies in NZ (23,24). Using a standardized computer-prompted protocol, participants were each interviewed by a trained interviewer about their food intake in the previous 24 h, including foods, beverages, and dietary supplements consumed both at and away from home. Interviews were conducted on weekdays and weekends using a multiple-pass technique with 4 stages. First, a ‘‘quick list’’ of foods and beverages consumed on the previous day was obtained. Detailed information of these foods, such as the brand and product name, cooking method, recipe used, time consumed, and the place the food was sourced, was collected in the second stage. In the third stage, the quantities of food consumed were estimated using food photographs, shape dimensions, food portion assessment aids, and packaging information. Finally, items were reviewed in chronological order to allow additions or changes to all aspects of the data. Food and beverages from the 24-h diet recalls were matched to the NZ Food Composition Database, or overseas food composition data when appropriate for nutrient estimation. Nutrient intakes were calculated without including dietary supplements (20).

household number, communication, and transport (26). NZDep2006 scores were assigned to participants based on their home addresses and divided into quintiles, in which quintile 1 represents participants with the lowest amount of deprivation (i.e., from the least-deprived areas) and quintile 5 represents participants with the most deprivation.

Results Participant characteristics. A total of 694 participants aged 15–18 y (325 males, 369 females) who completed the DHQ were included in the study. The mean 6 SE age was 16.5 6 0.1 y. Approximately 60% (n = 422, 54% females) of participants who had complete blood (n = 421) and urine (n = 411) variables were included in the biomarker subgroup analysis. There were no significant differences in the sex distribution, age, BMI, and HDHS-A scores between those excluded and included in the biomarker analyses (all P > 0.05; data not shown).

TABLE 1 Association between HDHS-A score and sociodemographic factors1

Total Sex Males Females Age (y) 15 16 17 18 Ethnicity NZEO Maori Pacific NZDep2006 quintile 1 (least deprived) 2 3 4 5 (most deprived) BMI categories3 Normal weight Overweight Obese

Participants

HDHS-A score

694

43.7 6 0.4

325 (47) 369 (53)

42.5 6 0.5a 45.0 6 0.5b

148 214 181 151

44.2 43.7 43.8 43.2

P2 ,0.001

0.84 (21) (31) (26) (22)

6 6 6 6

0.7 0.5 0.7 0.9 ,0.001

521 (75) 110 (16) 63 (9)

45.0 6 0.4a 39.8 6 0.7b 41.5 6 1.1b

133 155 123 136 147

45.5 44.9 44.2 43.1 40.6

,0.0014 (19) (22) (18) (20) (21)

6 6 6 6 6

0.9 0.5 0.9 0.8 0.7 0.52

394 (58) 171 (25) 85 (13)

43.5 6 0.4 43.8 6 0.7 45.2 6 1.2

Internal reliability of the HDHS-A. The items in the HDHS-A had low intercorrelations (0.10 # r # 0.13). The correlations between individual items with the total score were highest for item 9 [i.e., intake of potato and kumara (root vegetable) fries], followed by item 14 (i.e., soft drink or energy drink consumption) (Supplemental Table 2). The overall CronbachÕs a coefficient of the HDHS-A was 0.69, indicating that the index had good internal reliability (7).

1

HDHS-A scores vs. sociodemographic variables. The mean 6 SE HDHS-A was 44 6 0.4 (range, 16 to 66). HDHS-A score was associated with sex (P < 0.001), ethnicity (P < 0.001), and NZDep2006 quintile (P < 0.001) (Table 1). Females had significantly higher total HDHS-A scores than males. NZEO had higher HDHSA scores than Maori (P < 0.001) and Pacific (P < 0.01) groups.

in-depth and independent evaluation of the HDHS-A. Using a nationally representative sample and 2 independent methods for validation of the HDHS-A, the results from this study can therefore be generalized to NZ adolescents aged 15–18 y. For the first time, it was demonstrated by means of a diet index that the dietary habits of NZ adolescents differed by sex, ethnicity, and socioeconomic status. Compared with males, females scored higher in the HDHS-A, particularly in the following 3 clusters: 1) fat from meat, poultry, and fish; 2) other fats; and 3) fruit, vegetables, and bread (data not shown). This finding is in agreement with previous studies of adolescents that showed that females compared with males had intakes of fruit (35) and meat and meat products (36) that were more in line with dietary guidelines and better overall diet quality as measured by the Healthy Eating Index (14,37) and Mediterranean Dietary Score (38). The mean HDHS-A of the Maori and Pacific participants were 12% and 8%, respectively, lower than their NZEO counterparts. In addition, there was a trend for poorer dietary habits with increasing amount of deprivation (P-trend < 0.001). The associations of HDHS-A score with ethnicity and NZDep2006 were likely explained by the fact that a higher proportion of Maori and Pacific participants experience a higher amount of deprivation (P < 0.001; data not shown). In a previous regional study in Auckland, Maori and Pacific adolescents appeared to have less satisfactory dietary intakes as marked by higher fat intake and larger-than-standard portions of most food items but fewer daily servings of vegetables and cheese compared with NZ European and Asian adolescents (39). Considering dietary habits as a

HDHS-A score vs. nutrient intakes and nutritional biomarkers. The associations between HDHS-A scores and 24-h nutrient intakes and nutritional biomarker levels are shown in Table 2. Because HDHS-A score was negatively associated with energy intake (r = 20.18, P < 0.001), all nutrients were adjusted for total energy intake (in megajoules). Overall, higher relative intakes of protein, dietary fiber, PUFA, and lactose and lower intakes of sucrose were associated with increasing thirds of HDHS-A. Associations in the expected directions were also found with most micronutrient intakes, urinary sodium excretion, and whole-blood, serum, and RBC folate concentrations (P < 0.05).

Discussion This study describes the development and validation of a behavioral-based diet index, namely the HDHS-A, derived from self-reported dietary habits captured in a DHQ. The strength of this simple diet index lies in its practicality, because it is calculated based on dietary habits information that is easier to obtain than quantitative food and nutrient intake data. The use of 24-h diet recall and biomarker information allowed an

Values are n (%) or means 6 SEs weighted for the survey design. a,bSignificant differences between the groups when tested using post hoc linear combinations with BonferroniÕs adjustment (P , 0.017). HDHS-A, Healthy Dietary Habits Score for Adolescents; NZDep2006, 2006 New Zealand Deprivation Index; NZEO, New Zealand European and others. 2 Indicates level of significance in unadjusted regression models for the total sample. 3 As defined by Cole et al. (27,28). 4 Significant trend analysis.

Healthy dietary habits score for adolescents

939

Downloaded from jn.nutrition.org at UNIV OF CALIFORNIA SAN FRANCISCO KALMANOVITZ LIBRARY on January 17, 2015

scores would be found in participants with the following: 1) lower quintiles of NZDep2006; 2) more desirable nutrient profiles from the 24-h diet recall; and 3) more favorable blood and urinary biomarker profiles. Regression analyses were used to examine the associations between HDHS-A scores and sociodemographic variables (sex, age group, ethnicity, NZDep2006 quintile, BMI categories).When an association was significant, post hoc linear combinations with BonferroniÕs corrections were used to examine the differences among the groups. After categorizing total HDHS-A into thirds of low, medium, and high scores, ratios of nutrients-to-energy were calculated. A similar nutrient density approach for energy adjustment was adopted previously (33,34). Post hoc trend analyses were used to identify the presence of linear trend in nutrient intakes across the thirds of HDHS-A. To examine the association between nutritional biomarkers and HDHS-A score, trend analyses were conducted using biomarkers as the dependent variables and HDHSA thirds as the independent variables in the regression models.

TABLE 2

Adjusted nutrient intakes and nutritional biomarker levels by thirds of HDHS-A1 Thirds of HDHS-A Low 36 10 15 49 34 14 1.2 8.9 9.3 29 1.8 32 3.9 3.4 1.1 2.5 2.4 1.2 0.6 7.4 14 86 1.1 26 137 268 5.2 1.2 158 40 66 0.1 0.2 2.2 0.3 0.5 10 1.0 0.4 142 72 293 19 671

(35, 36) (9.5, 11.4) (14, 16) (48, 51) (33, 36) (14, 15) (1.2, 1.3) (8.4, 9.5) (8.8, 9.8) (28, 30) (1.7, 1.9) (27, 36) (3.6, 4.1) (3.2, 3.5) (1.1, 1.2) (2.2, 2.8) (2.1, 2.7) (1.0, 1.3) (0.5, 0.6) (6.4, 8.5) (13, 16) (76, 97) (1.1, 1.2) (25, 27) (127, 147) (254, 281) (4.6, 5.8) (1.1, 1.3) (138, 178) (35, 45) (60, 73) (0.1, 0.2) (0.2, 0.2) (1.9, 2.4) (0.2, 0.4) (0.4, 0.5) (8, 11) (0.9, 1.1) (0.4, 0.5) (131, 154) (67, 78) (268, 319) (17, 21) (617, 730)

45 9.6 16 49 34 14 1.3 9.3 9.1 29 1.9 28 3.8 3.3 1.2 2.4 2.2 1.5 0.4 7.0 14 87 1.2 29 143 295 5.6 1.2 236 36 76 0.2 0.2 1.9 0.2 0.4 10 1.0 0.5 112 68 306 22 689

(45, 45) (8.9, 10.3) (15, 16) (48, 51) (33, 36) (13, 15) (1.2, 1.4) (8.6, 10.0) (8.8, 9.5) (28, 30) (1.8, 2.1) (25, 30) (3.6, 3.9) (3.1, 3.5) (1.1, 1.2) (2.1, 2.6) (1.9, 2.4) (1.3, 1.7) (0.4, 0.5) (6.3, 7.8) (13, 15) (80, 95) (1.1, 1.3) (28, 30) (135, 152) (278, 313) (5.0, 6.3) (1.1, 1.3) (195, 277) (33, 39) (68, 83) (0.1, 0.2) (0.2, 0.2) (1.8, 2.1) (0.2, 0.2) (0.3, 0.4) (9, 12) (0.9, 1.0) (0.4, 0.6) (100, 125) (61, 76) (285, 330) (20, 24) (639, 744)

53 8.4 17 50 34 14 1.4 10.0 9.0 30 2.4 28 3.7 3.2 1.3 2.4 2.2 1.7 0.5 6.1 13 106 1.4 33 159 328 6.3 1.2 287 38 86 0.2 0.2 2.2 0.3 0.4 13 1.0 0.4 101 66 340 25 790

High

P-trend2

(53, 54) (7.9, 9.0) (16, 17) (48, 52) (32, 35) (13, 15) (1.3, 1.5) (9.6, 10.4) (8.6, 9.4) (28, 31) (2.2, 2.5) (25, 31) (3.4, 3.9) (3.0, 3.3) (1.2, 1.4) (2.2, 2.7) (2.0, 2.4) (1.5, 1.8) (0.5, 0.6) (5.5, 6.7) (12, 14) (98, 114) (1.2, 1.4) (32, 34) (154, 165) (313, 344) (5.7, 6.9) (1.2, 1.3) (251, 324) (34, 42) (80, 92) (0.2, 0.2) (0.2, 0.3) (2.0, 2.3) (0.2, 0.3) (0.4, 0.5) (11, 14) (1.0, 1.1) (0.4, 0.5) (90, 112) (59, 74) (312, 370) (22, 28) (722, 865)

,0.001 ,0.001 ,0.001 0.95 0.60 0.18 0.009 0.006 0.48 0.51 ,0.001 0.12 0.23 0.10 ,0.001 0.57 0.18 ,0.001 0.31 0.03 0.17 0.004 ,0.001 ,0.001 ,0.001 ,0.001 0.015 0.42 ,0.001 0.51 ,0.001 ,0.001 0.002 0.91 0.35 0.49 0.005 0.496 0.56 ,0.001 0.29 0.017 0.002 0.012

Values are means (95% CIs) weighted for the survey design. n = 694, participants who completed the dietary habits questionnaire; n = 422, participants who supplied a blood sample; n = 411, participants who supplied a urine sample. For all dietary variables, n = 255 for Low, n = 232 for Medium, and n = 207 for High categories. For all other variables, n differs for each parameter in the Low, Medium, and High categories. For biomarker variables (urinary iodine, urinary sodium, urinary potassium, whole-blood folate, serum folate, and RBC folate), values are geometric means (95% CIs). HDHS-A, Healthy Dietary Habits Score for Adolescents; U:S ratio, unsaturated fat (PUFA + MUFA)-to-SFA ratio. 2 Linear trend analysis using survey post-estimation. 1

whole by using a diet index approach, this study confirms findings of previous studies that disparities in diet quality of NZ adolescents are associated with ethnicity and socioeconomic status. This also provides evidence of the construct validity of the HDHS-A because it distinguishes differences in diet quality among ethnic and socioeconomic groups in NZ. HDHS-A scores were significantly associated with more favorable intake profiles of dietary fiber, PUFA, lactose, sucrose, 940

Wong et al.

calcium, iron, magnesium, phosphorus, potassium, selenium, and vitamins A, B-1 (thiamin), B-2 (riboflavin), and C. Of particular interest was the observed higher lactose but lower sucrose intakes. This finding is reflective of the index scoring system that rewards milk consumption but penalizes intakes of sugary foods, such as fruit juice, soft drinks, and confectionery (Supplemental Table 1). In addition, the HDHS-A seems to be a good indicator of types of fat in the diet. Although the total fat

Downloaded from jn.nutrition.org at UNIV OF CALIFORNIA SAN FRANCISCO KALMANOVITZ LIBRARY on January 17, 2015

HDHS-A Energy, MJ Protein, % energy Carbohydrate, % energy Fat, % energy SFA, % energy U:S ratio Protein, g/MJ Total fat, g/MJ Total carbohydrate, g/MJ Dietary fiber, g/MJ Cholesterol, g/MJ SFA, g/MJ MUFA, g/MJ PUFA, g/MJ Fructose, g/MJ Glucose, g/MJ Lactose, g/MJ Maltose, g/MJ Sucrose, g/MJ Total sugars, g/MJ Calcium, mg/MJ Iron, mg/MJ Magnesium, mg/MJ Phosphorus, mg/MJ Potassium, mg/MJ Selenium, mg/MJ Zinc, mg/MJ b-Carotene equivalents, mg/MJ Retinol, mg/MJ Total vitamin A, mg/MJ Thiamin, mg/MJ Riboflavin, mg/MJ Niacin, mg/MJ Vitamin B-6, mg/MJ Vitamin B-12, mg/MJ Vitamin C, mg/MJ Vitamin E, mg/MJ Urinary iodine, mmol/L Urinary sodium, mmol/L Urinary potassium, mmol/L Whole-blood folate, nmol/L Serum folate, nmol/L RBC folate, nmol/L

Medium

conform more closely to the NZ Food and Nutrition Guidelines for Healthy Children and Young People (21). However, it must be noted that the scoring of the HDHS-A remains arbitrary for some items, because not all dietary habits could be ordered on a scale according to their healthfulness. The scoring may be less sensitive in discerning subtleties in diet quality when a dietary habit was defined as neither optimum nor poor. Because of the cross-sectional nature of this study, reliability or temporal stability of the HDHS-A was not examined. Future work should examine the predictive capacity of the HDHS-A in relation to health outcomes in longitudinal studies. Contrary to most diet indices that use quantitative measures of nutrients and foods, the current index is based on 17 key dietary habits arranged in 5 dimensions or clusters. The HDHS-A measures both desirable and nondesirable dietary habits that are of nutritional concern in adolescence. The index also penalizes undesirable dietary habits relevant to youth culture, such as frequent intake of processed meat and soft drinks and purchasing food away from home (54,55). The significant associations between HDHS-A score and more favorable nutrient and biomarker profiles demonstrated the face and construct validity of the index. Overall, this study demonstrated that adoption of more healthful dietary habits, as measured by the HDHS-A, was associated with being female, not of Maori or Pacific ethnicity, and experiencing the least deprivation. The HDHS-A is valid as an indicator of diet quality because a higher diet score was associated with a more favorable nutrient and biomarker profiles. Derived from dietary habit questions, the HDHS-A is simple and practical to use and hence may be applied to small- and large-scale studies of adolescents in NZ, including those not collecting in-depth dietary intake data. Future studies are recommended to establish the predictive validity of HDHS-A in assessing health outcomes. Acknowledgments J.E.W., P.M.L.S., and W.R.P., designed the study; J.E.W. and S.M.W. analyzed the data; J.E.W. drafted the paper; and W.R.P. had responsibility for the final content. All authors read and approved the final manuscript.

Literature Cited 1.

2.

3.

4.

5. 6.

7.

8.

Deshmukh-Taskar PR, Nicklas TA, OÕNeil CE, Keast DR, Radcliffe JD, Cho S. The relationship of breakfast skipping and type of breakfast consumption with nutrient intake and weight status in children and adolescents: the National Health and Nutrition Examination Survey 1999–2006. J Am Diet Assoc. 2010;110:869–78. Rasmussen M, Krolner R, Klepp KI, Lytle L, Brug J, Bere E, Due P. Determinants of fruit and vegetable consumption among children and adolescents: a review of the literature. Part I: Quantitative studies. Int J Behav Nutr Phys Act. 2006;3:22. Sjo¨berg A, Hallberg L, Hoglund D, Hulthen L. Meal pattern, food choice, nutrient intake and lifestyle factors in The Goteborg Adolescence Study. Eur J Clin Nutr. 2003;57:1569–78. Mikkila¨ V, Rasanen L, Raitakari OT, Pietinen P, Viikari J. Consistent dietary patterns identified from childhood to adulthood: the cardiovascular risk in Young Finns Study. Br J Nutr. 2005;93:923–31. Hoelscher DM, Evans A, Parcel GS, Kelder SH. Designing effective nutrition interventions for adolescents. J Am Diet Assoc. 2002;102:S52–63. Buzzard IM, Stanton CA, Figueiredo M, Fries EA, Nicholson R, Hogan CJ, Danish SJ. Development and reproducibility of a brief food frequency questionnaire for assessing the fat, fiber, and fruit and vegetable intakes of rural adolescents. J Am Diet Assoc. 2001;101:1438–46. Magarey A, Golley R, Spurrier N, Goodwin E, Ong F. Reliability and validity of the ChildrenÕs Dietary Questionnaire; a new tool to measure childrenÕs dietary patterns. Int J Pediatr Obes. 2009;4:257–65. van Assema P, Brug J, Ronda G, Steenhuis I. The relative validity of a short Dutch questionnaire as a means to categorize adults and adolescents to total and saturated fat intake. J Hum Nutr Diet. 2001;14:377–90.

Healthy dietary habits score for adolescents

941

Downloaded from jn.nutrition.org at UNIV OF CALIFORNIA SAN FRANCISCO KALMANOVITZ LIBRARY on January 17, 2015

intake was constant within the acceptable macronutrient distribution range of 20–35% of total energy (40), the unsaturatedto-saturated fat ratio increased significantly across the thirds of HDHS-A (P < 0.001). This increasing fat ratio reflected a higher proportion of PUFA intake relative to SFA intake, i.e., a better fat quality across thirds of HDHS-A. Four studies examined a diet index in relation to nutritional biomarkers in children and adolescents (11,41–43). Recently, Vyncke et al. (42,43) studied the Flemish food-based Diet Quality Index for Adolescents in relation to a comprehensive range of blood biomarkers in a large multicenter cohort of European adolescents [the HELENA study (for Healthy Lifestyle in Europe by Nutrition in Adolescence)]. The Diet Quality Index for Adolescents was strongly associated with 2 nutritional biomarkers of long-term intake of vitamin D and vitamin B-12 and weakly associated with some serum FAs, such as EPA and DHA (42,43). However, they found no association with plasma folate (42), an indicator of acute folate status (44). In contrast, this study found significant associations between HDHS-A scores and 3 indices of folate status (i.e., serum, whole-blood, and RBC folate concentrations), suggesting that better shortand long-term folate status is associated with diet quality. This seems consistent with studies in adolescents and adults that also showed that a higher diet quality, using a diet index, was positively associated with either a higher serum folate (41,45) or both serum and RBC folate concentration (46). Given that discretionary use of salt cannot be accurately measured and the lack of a reliable local food composition database for dietary sodium analysis (20), urinary sodium was used as an indicator of sodium intake. Although dietary habits relating to sodium intake were not represented in the construct of the HDHS-A, a higher HDHS-A score was significantly associated with lower urinary sodium. This significant inverse association between a diet quality index and a biomarker of sodium has not been reported previously in the literature. Good habits with respect to sodium intake may also be practiced by participants with healthful dietary habits. Previous studies showed that health-promoting behaviors are likely to coexist (47,48), and our results may reflect such a phenomenon. There are several limitations to this study. A single 24-h diet recall provides a snapshot of dietary intake but does not represent the individualÕs habitual dietary intake, particularly for adolescents who are also known to have large within- and between-participant variability in intake (49). Nevertheless, single 24-h diet recalls are appropriate for estimating group means of the nutrient intake of a population when conducted within a random sample (50). As with other self-reported dietary methods, 24-diet recall data are susceptible to measurement errors related to forgetfulness and poor portion size estimation among adolescents (49,51,52). As part of a sensitivity analysis in this study, data were analyzed using a subsample with plausible energy reporting (n = 416), i.e., those with a ratio of total energy intake-to-basal metabolic rate of >1.1 (53). After excluding energy underreporters, the associations between HDHS-A scores and nutrients remained significant for most nutrients, except for lactose, selenium, riboflavin, niacin, and vitamin C. In addition, significant inverse associations were found for total carbohydrate (P-trend < 0.05), SFA (P-trend < 0.05), and total sugars (P-trend < 0.001) across the thirds of HDHS-A. To preserve the representativeness of the survey data, possible underreporters were included in the current analysis. Therefore, the results presented represent a conservative estimate of the validity of the HDHS-A. The scoring system of the HDHS-A may have implications for its usefulness. Higher scores were assigned to dietary habits that

9. 10.

11.

12.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29. 30.

31.

942

Wong et al.

32. Pino S, Fang SL, Braverman LE. Ammonium persulfate: a safe alternative oxidizing reagent for measuring urinary iodine. Clin Chem. 1996;42:239–43. 33. Brown CC, Kipnis V, Freedman LS, Hartman AM, Schatzkin A, Wacholder S. Energy adjustment methods for nutritional epidemiology: the effect of categorization. Am J Epidemiol. 1994;139:323–38. 34. Kipnis V. Structure of dietary measurement error: results of the OPEN Biomarker Study. Am J Epidemiol. 2003;158:14–21. 35. Savige GS, Ball K, Worsley A, Crawford D. Food intake patterns among Australian adolescents. Asia Pac J Clin Nutr. 2007;16:738–47. 36. Kobe H, Stimec M, Ribic CH, Fidler Mis N. Food intake in Slovenian adolescents and adherence to the Optimized Mixed Diet: a nationally representative study. Public Health Nutr. 2012;15:600–8. 37. de Andrade SC, de Azevedo Barros MB, Carandina L, Goldbaum M, Cesar CL, Fisberg RM. Dietary quality index and associated factors among adolescents of the state of Sao Paulo, Brazil. J Pediatr. 2010;156:456–60. 38. Mart´ınez E, Llull R, Del Mar Bibiloni M, Pons A, Tur JA. Adherence to the Mediterranean dietary pattern among Balearic Islands adolescents. Br J Nutr. 2010;103:1657–64. 39. Sluyter JD, Schaaf D, Metcalf PA, Scragg RK. Dietary intakes of Pacific, Maori, Asian and European adolescents: the Auckland High School Heart Survey. Aust N Z J Public Health. 2010;34:32–7. 40. National Health and Medical Research Council. Nutrient reference values for Australia and New Zealand including recommended dietary intakes. Canberra, Australia: National Health and Medical Research Council; 2006. 41. Kleiser C, Mensink GB, Scheidt-Nave C, Kurth BM. HuSKY: a healthy nutrition score based on food intake of children and adolescents in Germany. Br J Nutr. 2009;102:610–8. 42. Vyncke K, Cruz Fernandez E, Fajo-Pascual M, Cuenca-Garcia M, De Keyzer W, Gonzalez-Gross M, Moreno LA, Beghin L, Breidenassel C, Kersting, M, et al. Validation of the Diet Quality Index for Adolescents by comparison with biomarkers, nutrient and food intakes: the HELENA study. Br J Nutr. 2013;109:2067–78. 43. Vyncke KE, Huybrechts I, Dallongeville J, Mouratidou T, Van Winckel MA, Cuenca-Garcia M, Ottevaere C, Gonzalez-Gross M, Moreno LA, Kafatos AG, et al. Intake and serum profile of fatty acids are weakly correlated with global dietary quality in European adolescents. Nutrition. 2013;29:411–9. 44. Gibson RS. Reproducibility in dietary assessment. Principles of Nutritional Assessment. 2nd ed. New York: Oxford University Press; 2005:129–48. 45. Kant AK, Graubard BI. A comparison of three dietary pattern indexes for predicting biomarkers of diet and disease. J Am Coll Nutr. 2005;24:294–303. 46. Weinstein SJ, Vogt TM, Gerrior SA. Healthy Eating Index scores are associated with blood nutrient concentrations in the third National Health and Nutrition Examination Survey. J Am Diet Assoc. 2004;104:576–84. 47. Ottevaere C, Huybrechts I, Benser J, De Bourdeaudhuij I, Cuenca-Garcia M, Dallongeville J, Zaccaria M, Gottrand F, Kersting M, Rey-Lopez JP, ´ et al. Clustering patterns of physical activity, sedentary and dietary behavior among European adolescents: the HELENA study. BMC Public Health. 2011;11:328. 48. Sanchez A, Norman GJ, Sallis JF, Calfas KJ, Cella J, Patrick K. Patterns and correlates of physical activity and nutrition behaviors in adolescents. Am J Prev Med. 2007;32:124–30. 49. Livingstone MB, Robson PJ. Measurement of dietary intake in children. Proc Nutr Soc. 2000;59:279–93. 50. McPherson RS, Hoelscher DM, Alexander M, Scanlon KS, Serdula MK. Dietary assessment methods among school-aged children: validity and reliability. Prev Med. 2000;31:S11–33. 51. Collins CE, Watson J, Burrows T. Measuring dietary intake in children and adolescents in the context of overweight and obesity. Int J Obes (Lond). 2010;34:1103–15. 52. Livingstone MB, Robson PJ, Wallace JM. Issues in dietary intake assessment of children and adolescents. Br J Nutr. 2004;92(Suppl 2):S213–22. 53. Goldberg GR, Black AE, Jebb SA, Cole TJ, Murgatroyd PR, Coward WA, Prentice AM. Critical evaluation of energy intake data using fundamental principles of energy physiology: 1. Derivation of cut-off limits to identify under-recording. Eur J Clin Nutr. 1991;45:569–81. 54. Gillis LJ, Bar-Or O. Food away from home, sugar-sweetened drink consumption and juvenile obesity. J Am Coll Nutr. 2003;22:539–45. 55. Taveras EM, Berkey CS, Rifas-Shiman SL, Ludwig DS, Rockett HR, Field AE, Colditz GA, Gillman MW. Association of consumption of fried food away from home with body mass index and diet quality in older children and adolescents. Pediatrics. 2005;116:e518–24.

Downloaded from jn.nutrition.org at UNIV OF CALIFORNIA SAN FRANCISCO KALMANOVITZ LIBRARY on January 17, 2015

13.

Yaroch AL, Resnicow K, Khan LK. Validity and reliability of qualitative dietary fat index questionnaires: a review. J Am Diet Assoc. 2000;100:240–4. Kontogianni MD, Farmaki AE, Vidra N, Sofrona S, Magkanari F, Yannakoulia M. Associations between lifestyle patterns and body mass index in a sample of Greek children and adolescents. J Am Diet Assoc. 2010;110:215–21. Chiplonkar SA, Tupe R. Development of a diet quality index with special reference to micronutrient adequacy for adolescent girls consuming a lacto-vegetarian diet. J Am Diet Assoc. 2010;110:926–31. Farajian P, Risvas G, Karasouli K, Pounis GD, Kastorini CM, Panagiotakos DB, Zampelas A. Very high childhood obesity prevalence and low adherence rates to the Mediterranean diet in Greek children: the GRECO study. Atherosclerosis. 2011;217:525–30. Feskanich D, Rockett HR, Colditz GA. Modifying the Healthy Eating Index to assess diet quality in children and adolescents. J Am Diet Assoc. 2004;104:1375–83. Hurley KM, Oberlander SE, Merry BC, Wrobleski MM, Klassen AC, Black MM. The healthy eating index and youth healthy eating index are unique, nonredundant measures of diet quality among low-income, African American adolescents. J Nutr. 2009;139:359–64. Lazarou C, Panagiotakos DB, Matalas AL. Level of adherence to the Mediterranean diet among children from Cyprus: the CYKIDS study. Public Health Nutr. 2009;12:991–1000. Serra-Majem L, Ribas L, Garcia A, Perez-Rodrigo C, Aranceta J. Nutrient adequacy and Mediterranean Diet in Spanish school children and adolescents. Eur J Clin Nutr. 2003;57(Suppl 1):S35–9. Kosti RI, Panagiotakos DB, Mariolis A, Zampelas A, Athanasopoulos P, Tountas Y. The Diet-Lifestyle Index evaluating the quality of eating and lifestyle behaviours in relation to the prevalence of overweight/obesity in adolescents. Int J Food Sci Nutr. 2009;60(Suppl 3):34–47. Lazarou C, Panagiotakos DB, Spanoudis G, Matalas AL. E-KINDEX: a dietary screening tool to assess childrenÕs obesogenic dietary habits. J Am Coll Nutr. 2011;30:100–12. Turconi G, Guarcello M, Maccarini L, Cignoli F, Setti S, Bazzano R, Roggi C. Eating habits and behaviors, physical activity, nutritional and food safety knowledge and beliefs in an adolescent Italian population. J Am Coll Nutr. 2008;27:31–43. University of Otago, Ministry of Health. Methodology report for the 2008/09 New Zealand Adult Nutrition Survey. Wellington, New Zealand: Ministry of Health; 2011. Ministry of Health. Food and Nutrition Guidelines for Healthy Children and Young People (Aged 2–18 Years). A background paper. Wellington, New Zealand: Ministry of Health; 2012. Ministry of Health. Healthy Eating—Healthy Action Oranga Kai— Oranga Pumau: a strategic framework. Wellington, New Zealand: Ministry of Health; 2003. Russell DG, Parnell WR, Wilson NC. NZ Food: NZ People. Key results of the 1997 National Nutrition Survey. Wellington, New Zealand: Ministry of Health; 1999. Clinical Trials Research Unit, Synovate. A national survey of children and young peopleÕs physical activity and dietary behaviours in New Zealand: 2008/09. Key findings. Wellington, New Zealand: Ministry of Health; 2010. Statistics New Zealand. 2006 Census Data—QuickStats about New ZealandÕs population and dwellings—tables. 2006 [cited 2013 Dec]. Available from: http://www.stats.govt.nz/Census/2006CensusHomePage/ QuickStats/quickstats-about-a-subject/nzs-population-and-dwellings.aspx. Salmond C, Crampton P, King P, Waldegrave C. NZDep: a New Zealand index of socioeconomic deprivation for individuals. Soc Sci Med. 2006;62: 1474–85. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320:1240–3. Cole TJ, Flegal KM, Nicholls D, Jackson AA. Body mass index cut offs to define thinness in children and adolescents: international survey. BMJ. 2007;335:194. OÕBroin S, Kelleher B. Microbiological assay on microtitre plates of folate in serum and red cells. J Clin Pathol. 1992;45:344–7. Senti FR, Pilch SM. Analysis of folate data from the second National Health and Nutrition Examination Survey (NHANES II). J Nutr. 1985;115:1398–402. Institute of Medicine. Dietary reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline. Washington: National Academies Press; 1998.