Diet quality is associated with obesity and hypertension in Australian adults: a cross sectional study

Body weight (BW; kg), height (cm) and waist circumference (WC; cm) measurements were measured on a voluntary basis by trained interviewers using digital scales, a stadiometer and a metal tape respectively. Pregnant women were not measured. Subjects were encouraged to remove their shoes and any heavy clothing prior to having measurements taken. Body mass index (BMI) was derived using Quetelet’s metric (kg/m²). Standard cut offs for BMI and WC were applied: underweight/normal weight: BMI < 25 kg/m²; overweight: BMI ≥ 25 kg/m² and <30 kg/m²; obese: BMI ≥ 30 kg/m²; and central adiposity: WC > 102 cm (men) and >88 cm (women) [16].

Blood pressure measurements were voluntary. Systolic (SBP) and diastolic blood pressure (DBP) measurements were taken on the left arm, unless there was a prohibitive reason such as an injury. Interviewers undertook two blood pressure readings using an automated blood pressure monitor in which systolic and diastolic pressures were displayed. Individuals were categorized as non-hypertensive (blood pressure <140/90 mmHg) and hypertensive (≥140/90 mmHg). Data on hypertensive medication were not recorded [15].

Socio-demographic characteristics were collected in the NNPAS via interviewer-administered questionnaires. Smoking was defined as ex-smoker, current smoker and never smoked. Education status was operationalized as low (some high-school or less), medium (high-school or some high-school and/or certificate/diploma) and high (tertiary qualification) [15]. Urban or rural location was defined as major city, inner rural or other [15]. Physical activity (PA) was assessed according to i) meeting PA guidelines (150 min of PA per week and 150 min of PA over 5 or more sessions per week) and ii) time spent sedentary (minutes per day). Female life stage was operationalized as: never having menstruated, experiencing menopause or post-menopause. Further details are provided elsewhere [15].

An automated, multiple-pass, 24-h dietary recall was used to provide quantitative information on foods and beverages consumed on the day prior to interview based on the Agricultural Research Service of the United States Department of Agriculture Automated Multiple-Pass Method [17]. The interview was divided into five phases: quick list description of food and beverages consumed the previous day, prompt to remember any omitted foods, information on time and eating occasion, further details and a final probe to recall any omitted foods of beverages [15]. A second 24-h recall, via telephone interview, was collected at least 8 days after the first interview. Nutrient intakes were derived from the 24-h recalls using the Australian Supplement and Nutrient Database 2011–13, developed by Food Standards Australia New Zealand [18]. Information on type of milk consumed, usual daily intake of fruit and vegetables and use of salt were collected in the NNPAS survey [15]. Energy misreporting was calculated as the ratio of energy intake to predicted total energy expenditure (using sex and age-specific equations for a range of weight status, assuming a PA level of "low active" PA level ≥ 1.4 and PA level < 1.6) [19].

The DGI is a food-based score designed to reflect the diet quality of subjects according to compliance with the 2013 ADG for Australian adults [20]. The DGI used in the present study was based the DGI-2013 [21], and thus comprised a total of 23 items (see Additional file 1: Table S1), but was adapted from use in food frequency questionnaires (FFQ) to use in the present 24-h recalls. DGI scores ranged between 0 and 130, with a higher score indicating better diet quality. Each item was scored out of ten (zero indicating the guideline was not met). Items with two sub-components were scored out of five. Cut-offs used to obtain the maximum score for each component were tailored to age and sex-specific food-based recommendations outlined in the ADG [22]. Proportionate scores were derived where intakes fell between the maximum and minimum soring criteria for all items except discretionary foods, saturated and unsaturated fat, salt, sugar and alcohol [23, 24].

A food variety score was estimated based on the variety of foods consumed within the five core food groups: fruit, vegetables, meat or meat alternatives, dairy and cereals [25]. Foods scored 1 point if consumed above a cut off (>15 g/d for beverages and >20 g/d for foods), analogous to the RFS [26], and variety was estimated within each core food group by summing scores for each food group and dividing by the total number of foods within each core food group. Scores were summed across the five core food groups and multiplied by two to create a score out of 10 [27]. Salt intake was assessed using two questions from the food habits and attitudes questionnaire: frequency of adding salt during cooking and during a meal and saturated fat intake was scored against the ratio of trimmed meat to total meat intake and low fat milk to total milk intake. Sub-components of cereal, meat and alternatives and fluid intake captured information on the ratio of wholegrain bread to total bread, lean meat to total meat and water to total fluid intake, respectively. Total beverage intake included milk and soy beverages, smoothies, fruit and vegetable juices, low calorie cordials, low calorie soft drinks, water, tea and coffee. Flavored milk drinks, fruit drinks, high-sugar soft drinks, cordials and alcohol were not included in total beverage intake due to the associations with dental caries, weight gain and diabetes [28], but were included as “Discretionary foods”. “Discretionary foods”, defined as energy-dense foods and drinks that are not essential to nutrition, included sugar-sweetened beverages, sweet and salty snacks and confectionary, cakes and pastries, high-fat processed meats and dishes, pies, fried foods, ice cream and other dairy desserts, cream, butter and spreads and alcoholic beverages [29]. Cut-offs for discretionary food intakes were sourced from the ADG companion resource for educators [22]. Unsaturated fats included intakes of nuts, seeds and margarine, while added sugar included confectionary, jam, marmalade, honey, syrup and sugar-sweetened beverages.

The RFS is a food-based diet variety score calculated based on the frequency of consumption of foods from five core food groups: fruits (6 items), vegetables (9 items), wholegrains (4 items), lean meats and alternatives (2 items) and low-fat dairy (2 items). Scores ranged between 0 and 23, with a higher score indicating a better diet quality. Scoring was based on a method by Kant and Graubaud [30] for use with 24-h recall data, where foods were assigned a score of 1 if they were consumed above the minimum amount threshold: 15 g/d for non-beverages and 30 g/d for beverages. Two additional methods for calculated RFS were tested. The first of these methods assigned a score of 1 for each recommended food if consumption was ≥0.5 servings over the 2 day recalls [27]. The second additional method was based on sex-specific median cut offs for consumers, where a score of 1 was allocated if consumption was above the median cut off [31].

Participants were excluded from the present analyses if they i) were pregnant and/or breastfeeding ii) had missing data for outcomes and covariates ii) only 1 day of dietary recall. All analyses were conducted for men and women separately. Variables were tested for skewness and kurtosis and were log transformed if not normally distributed. BW, BMI, and WC were log transformed prior to analysis. Tertiles of diet quality score were selected as the optimum methodology for evaluating variations in diet quality score based on maximizing power from the sample size and previous literature [32]. Linear regression and chi squared tests were used to test for significant differences in participant characteristics across tertile of diet quality score and in diet quality score between sexes for categorical and continuous variables respectively. To answer our primary research question (Does the odds ratio of obesity, central adiposity and hypertension vary by tertile of diet quality score?), multi-variable-adjusted logistic regression analyses were performed. To assess these relationships in continuous outcomes multiple linear regression analyses were performed. Analyses were adjusted for age (continuous), smoking (categorical), physical activity (whether met PA recommendations; binary), education (categorical), urban or rural location (categorical), energy misreporting (ratio of energy intake to predicted total energy expenditure; continuous), dieting or atypical dietary intake on day of reporting (categorical) and female life stage (categorical; women only). Hypertension-related outcomes were further adjusted for BMI (continuous). To answer our secondary research question (Does the odds ratio of hypertension by tertile of diet quality score vary according to obesity status of the population?), multi-variable-adjusted logistic regression analyses stratified by BMI status (normal weight vs. overweight or obese) and by central adiposity (no central adiposity vs. central adiposity) were performed. Analyses were adjusted for the same covariates mentioned above with the exception of BMI.

Data were analyzed using Stata (version 14; StataCorp., College Station, TX, USA) using survey weightings for analyzing complex survey data to account for the survey design. These weightings were specifically designed to account for bias associated with those who volunteered to complete the second day of dietary recalls (64 %; n = 7735). Weighting was calibrated to align with independent benchmarks in designated categories of sex by age and area of usual residence [15]. P < 0.05 was considered statistically significant.