Challenges and barriers to healthy eating habits and food choices : a discrete choice study

Abstract

Aim of the study: This study aimed to distinguish specific food product characteristics that lead to healthy food product choices in two cohorts of different educational levels. Preferred food product characteristics and influencing factors were linked, namely, price, healthiness, and taste, impacting food choices according to participants’ sociodemographic characteristics. Methods: A total of 406 participants from different educational levels were recruited, through a mixture of sampling techniques, and took part in an online discrete choice experiment (DCE) that elicited participant stated preferences (SP) for food choice questions (n = 10). Participants were presented with equal amounts of two (2) food items of the same food source (for example: one kilogram (1kg) of chicken breast and one kilogram (1kg) of chicken nuggets). Each choice task contained four (4) pairs of different prices, and for each price-pair, participants were asked to choose one price from each pair. This descriptively captured respondents’ willingness to pay for one product over another, whilst is elicited the participants’ trade off point (if any) for the selected food products, based on the price. Participant sociodemographic data was also captured at the beginning of the DCE questionnaire. Data was exported from Qualtricsxm as an IBM SPSS statistical software package compatible file, and data analysis was carried out on SPSS. Chisquared test of association was used to evaluate associations amongst and between variables and all p values generated were statistically significant if they resulted less than 0.05. ANOVA and T-tests were used to derive unadjusted means and standard errors. P-values were also generated to assess the statistical significance between each food product characteristic (e.g., as tastiness and healthiness) and demographic variables. Mental health data was analysed using the Mental Health Inventory-5 (MHI-5) score. Multivariate analysis was performed to minimise confounding effects and link the true significance of these variables. A forward stepwise linear regression was performed for the most significant demographic variables on the food product characteristics and mental health. This study received ethical approval from University of Malta’s Faculty Research Ethics Committee (FREC) prior to its initiation. Results: Most of the data was collected from Maltese citizens of a healthy weight, or who were in the overweight BMI category. Results for the DCE choice tasks showed that educational level had minimal effect on healthy food product choices, and that healthy weight individuals were more likely to give more value to food product healthiness over price. A link was also established between female respondents and price-driven choices. Discussion and conclusions: The analysis revealed further complexities with certain food products, which was most prominent during the analysis and results presentation of the second choice task in the DCE tool (whole/ skimmed milk), where anomalous results were a consequence an interplay of attributes that went beyond healthiness and price. This could have also indicated that other underlying social and contextual factors might have been affecting food choices, clearly impacting the respondents’ decision-making processes. One of the most significant links that were established in this study revealed how female participants were more price-conscious when making food choices over male participants. From the multivariate analysis of food attributes’ importance to respondents, most sociodemographic characteristics transpired to have significant importance, however educational level was never significant.