An empirical DP predominated by foods high in sugar and fibre, and low in DED and fat was identified among the study adolescents in the PUTRA-Adol follow-up study in 2019–2020. Moreover, female adolescents showed significant higher overall life satisfaction score compared to male adolescents in the PUTRA-Adol follow-up study in 2019–2020. A significant prospective association between adherence to the identified DP z-scores and school domain, in male adolescents and an inverse relationship between the DP z-score and self-domain in all adolescents was observed over the study period. These associations were still apparent even after adjusting for a wide range of covariates including race, school location, maternal education and baseline dietary misreporting.

The identified ‘High sugar and fibre, and Low DED and fat’ DP in the 2019–2020 PUTRA-Adol follow-up study was slightly different to what we found in the baseline study in 2016 [12]. The main observed difference was of that correlation between DED and the identified DPs at both time points (r = 0.39 in 2016 vs. r = 0.01 in 2019–2020). Nonetheless, the low correlations between DED and the identified pattern in the current study was somewhat not surprising. This is mainly due to the changes in the eating behaviour which were rather expected during adolescence and the fact that the follow-up study was conducted in the early stage of COVID-19 pandemic when access to foods high in DED might have been restricted. The consumption of foods low in DED were probably compensated by foods high in sugar such as SSB in the follow-up study. Having said that, the contributions of key food groups loaded on both DPs identified in the baseline study in 2016 and follow-up study in 2019–2020 were largely similar [10]. These key groups were including SSBs and sweets and were found to be similar to those in the UK ALSPAC study among children and adolescents aged between 7 and 15 years of age and in the Raine study in Australia among adolescents aged between 13 and 17 years of age [28, 29]. Higher adherence to DPs highly loaded on SSBs and sweets including chocolate and confectioneries were found to be prospectively associated with adiposity [28] and other cardiometabolic risk factors [12, 29]. Findings on life satisfaction in the follow-up PUTRA-Adol follow-up study were similar to that baseline study conducted in 2016 where female adolescents generally reported a higher overall life satisfaction score [10]. Female adolescents showed higher scores in all other domains including family, school, friends and living environment domains in the baseline study in 2016 compared to male adolescents who only had higher scores in the self-domain [10]. However, in the current PUTRA-Adol follow-up study male adolescents showed higher score in family, school and living environment compared to the female adolescents who only showed higher scores for friends and self-domains. This finding might indicate a potential difference in social connections for male and female adolescents, where male adolescents tend to find greater attachment and satisfaction within their home environment as they grow while female adolescents appeared to prioritize value on friendships.

A greater adherence to the identified dietary patterns was prospectively associated with increasing life satisfaction score for the school domain between 13 and 17 years of age could be due to the connections and interactions with like-minded peers as well as exposure and consumption to foods and drinks that are highly palatable and pleasurable may have contributed to adolescents' high levels of life satisfaction in the school vicinity. On the other hand, our study found an increasing dietary pattern score was associated with a decreasing score for self-domain for overall adolescents. In the current follow-up study, we found that one-third of the adolescents disagreed that they look good. Since life satisfaction with the overall appearance is linked to self-esteem, any dissatisfaction with oneself might have largely contributed to the low self-esteem and subsequently unhealthy eating behaviours particularly intakes of foods high in sugar and fat [30].

To the best of knowledge, no studies have assessed associations between empirically derived adolescent DPs and life satisfaction per se except to that published in the baseline study in 2016 [10]. Previously, we found a significant cross-sectional association between adherence to the ‘High sugar, High fibre, High DED and Low fat’ DP and overall life satisfaction and living environment domain, in female adolescents, in the baseline study [10]. The fact that a slightly different DP (‘High sugar, High fibre, Low DED and Low fat’) was observed in the current PUTRA-Adol follow-up study, this would have led to the differences in findings between the DPs and overall life satisfaction. Secondly, SSBs was highly correspond to the identified DP in the baseline study in 2016 and that has changed to sweets in 2019–2020. Since liquid SSBs are high in added-sugar content this could have led to low satiety and incomplete compensatory reduction in energy intake at subsequent meals after consuming them, as opposed to sweets as in solid food form. Nonetheless, the contribution of free sugar in the PUTRA-Adol follow-up study is mainly from the sweets and SSB which represents the characteristic of a ‘western’ DP and is most likely linked to the observed connection between DPs and lower life satisfaction. Correspondingly, a review in 2011 reported that ‘Western’ DP characterised by foods and beverages with high glycemic loads such as SSB, sweets, candies, and confectionaries were found to have raised plasma C-reactive protein (CRP) which is a marker for inflammation and subsequently depression, anxiety, fatigue, and social withdrawal in young people [28, 29, 31]. Furthermore, a Western Australian study found a prospective association between adherence to a "Western" DP characterised by confectionaries and mental illnesses during adolescence [9]. Moreover, a prospective study of Australian adolescents found that a lower Paediatric Quality of Life Inventory (PedsQL) score was associated with lower diet quality, which was measured as lower adherence to the Dietary Guidelines for Children and Adolescents in Australia and a greater consumption of processed food including those confectionaries [31]. Besides that, Supplémentation en Vitamines et Minéraux Antioxydants (SU.VI.MAX) cohort study among children found an association between junk food in early childhood and emotional problems that may reflect a long term nutritional imbalance [33]. Declines in psychological functioning were also linked to lower diet quality characterised by high glycemic load (GL) diets over the course of a follow-up period among Australian adolescents [32]. Whereas, Mediterranean diet which includes high number of healthy foods such as whole grains, fruits, vegetables, seafood, and nuts showed a lower risk for mental health among adolescents in the SU.VI.MAX cohort study [33].

The strength of this study included the ability to track dietary habits during adolescence, an age group that was often overlooked in the previous years. It is also important to note that this was the first study to investigate the prospective association between DP and life satisfaction level among Malaysian adolescents. There are several limitations worth mentioning. Firstly, the large reduction in the response rate in the present follow-up study compared to the baseline study. This was mainly due to the Covid-19 pandemic issue, when all schools in Malaysia were closed and travels were hindered for data collection. While we attempted for online data collection, a prevalent issue in the majority of households stemmed from a lack of internet access, consequently posing further hurdles to the data collection process. Furthermore, the potential for reporting bias was acknowledged due to the self-administration of the FFQ and MSLSS. To mitigate this bias, researchers offered illustrative examples and meticulously reviewed the questionnaires for any discrepancies. Moreover, the associations observed in this study could be a result of reverse causation. In this case, low life satisfaction may be linked to increased sugar intake as mental health problems such as depression could negatively impact dietary choices. Additionally, regression models were also adjusted to accommodate dietary misreporting, a prevalent concern within the field of nutrition research. Although adjustment for a few confounders was done, residual confounding cannot be fully ruled out as there were important additional confounding factors such as BMI and physical activity that were not considered in this study. Furthermore, the study participants only came from three southern states in Malaysia, therefore the findings may not be generalisable to the entire country.