A new study has found that individuals who prefer sweet foods and drinks are more likely to experience depression. Published in the Journal of Translational Medicine, the research revealed that people with a “sweet tooth” had a 31% higher likelihood of depression compared to those who preferred healthier options.
The new study also sought to address gaps in nutrition research by using data-driven methods to explore the biological pathways that could connect dietary habits to mental health and other health conditions.
“In the field of nutrition research, there is still an unmet need for clarity of the effects of food types we choose on our health,” said study author Hana Navratilova, a PhD candidate at the University of Surrey. “This can be addressed by leveraging data-driven methods that offer practical solutions for nutrition problems. Developments in this area offer clear benefits for nutritionists, healthcare professionals, as well as clients/patients. For example, a nutritionist can get a gist of a client’s health risk and, by focusing on clients’ food preferences, tailor nutritional advice more effectively. From a client’s perspective, they can identify their risks before consulting a nutritionist or dietitian for further advice.”
The study analyzed data from the UK Biobank, a large biomedical database that includes health and lifestyle information from over 500,000 participants aged 40 to 69 years. For this research, the team focused on 180,000 individuals who completed a detailed Food Preference Questionnaire in 2019. The questionnaire asked participants to rate their preference for 140 food items—such as fruits, vegetables, meats, sweets, and beverages—on a nine-point scale ranging from extreme dislike to extreme like. The study excluded participants who had a significant number of missing or incomplete responses to ensure accuracy.
The researchers employed latent profile analysis, a statistical method, to categorize participants into three distinct food preference profiles based on their questionnaire responses: (1) Health-conscious, who favored fruits, vegetables, and healthier food options, (2) Omnivores, who enjoyed a broad range of foods, including meats, fish, and some sweets, and (3) Sweet tooth, who had a strong preference for sugary foods and drinks and a lower interest in healthier foods like fruits and vegetables.
“Similar to an MBTI (Myers–Briggs Type Indicator) personality test, our study provides an overview of someone’s health profile based on their food preferences,” Navratilova told PsyPost. “However, this profile isn’t fixed, it helps identify areas for improvement to achieve better health.
In addition to the food preference data, the researchers analyzed health outcomes and biological markers in blood samples from the participants. They examined how each food preference profile was associated with conditions such as depression, diabetes, stroke, and other chronic diseases. They also examined biomarkers such as blood sugar, cholesterol, and inflammatory markers to assess any metabolic differences between the groups.
The study’s key finding was that individuals in the sweet tooth group were at significantly higher risk for various health issues. Specifically, they were 31% more likely to suffer from depression compared to those in the other two groups. Additionally, participants with a strong preference for sweet foods were more likely to have higher rates of diabetes and stroke. These findings suggest that a diet high in sugary foods and drinks may negatively impact both mental and physical health.
In contrast, the health-conscious group, which preferred fruits and vegetables, exhibited more favorable health outcomes. They had lower levels of inflammatory markers, healthier cholesterol profiles, and were at reduced risk for conditions like diabetes and cardiovascular diseases. The omnivore group fell between the health-conscious and sweet tooth groups in terms of health risks, reflecting their more balanced but less targeted dietary choices.
The biological analysis revealed notable differences in blood markers between the groups. For example, the health-conscious group had higher levels of beneficial fatty acids and ketone bodies, which are associated with better metabolic health. The sweet tooth group, on the other hand, exhibited higher levels of biomarkers linked to poorer metabolic outcomes, such as elevated blood glucose. These metabolic differences provide insight into how dietary preferences might contribute to the development of chronic diseases over time.
“What’s surprising is that these profiles were identified without relying on actual food intake, only individual preferences as reported by over 180,000 people in the UK Biobank, yet we were able to identify biomarkers that are meaningful to health status,” Navratilova said.
“Our findings show that essentially, you are what you like to eat,” added senior author Nophar Geifman, a professor of health and biomedical informatics at the University of Surrey. “But, we don’t want people to walk away from reading this research thinking their future health is pre-determined and fixed, that whichever foods they tend to like will directly affect their health outcomes. There is an obvious link between what we like to eat and what we actually eat – but individuals have a choice; Increasing dietary fiber intake while reducing that of sugars and ultra-processed foods will contribute to better health outcomes.”
Although this study offers valuable insights, it is not without limitations. One of the main limitations is that it relied on self-reported data from participants, which can introduce bias. People may not always accurately report their food preferences or health conditions. Additionally, the cross-sectional nature of the study means that it cannot establish cause and effect—only associations between food preferences and health risks.
“These findings still need to be validated in different cohorts and populations to ensure their wider generalizability,” Navratilova said.
For future research, the authors suggest exploring whether personalized nutrition advice based on food preferences could help reduce the risk of chronic diseases. They also recommend further studies that track changes in food preferences over time and how these changes might influence health outcomes. The research team aims to develop tools that could help individuals, nutritionists, and healthcare providers use data about food preferences to offer more personalized and effective dietary advice.
“Building on this work, our longer-term goal is to develop AI-based personalized nutrition tools that can enable individuals, nutritionists and healthcare providers, to make informed decisions about their diet and health that are also aligned with their own personal circumstances and preferences,” Navratilova said.
The study, “Artificial intelligence driven definition of food preference endotypes in UK Biobank volunteers is associated with distinctive health outcomes and blood based metabolomic and proteomic profiles,” was authored by Hana F. Navratilova, Anthony D. Whetton, and Nophar Geifman.
