The analysis identified four subgroups of eaters: unhealthy eaters, passive eaters, conscious eaters, and balanced eaters. The labels describe more than dietary preference; they capture entire orientations toward food, from deliberate attention to nutrition and food labels to habitual consumption of items high in fat, sugar, or sodium, often abbreviated as HFSS foods. The existence of these distinct classes matters because public health interventions tend to be designed for average populations, while actual behavior is patterned. A campaign that resonates with a conscious eater may fall flat with a passive eater, and the study’s statistical approach makes those differences visible in a way that simple averages cannot.

Latent class analysis works by modeling the probability that individuals share an underlying, unobserved categorical trait based on their observed responses. In this study, the researchers compared competing models with different numbers of classes using standard fit criteria, including the Akaike information criterion, the Bayesian information criterion, the consistent Akaike information criterion, and the sample size-adjusted Bayesian information criterion. These indices penalize model complexity while rewarding explanatory power, allowing the team to settle on the four-class solution as the best balance between parsimony and accuracy. Once the classes were established, binary logistic regression was used to estimate the odds of belonging to particular subgroups based on sociodemographic characteristics and indicators of health lifestyle.

The setting is significant. Universities concentrate diverse groups of students and working adults in one environment, making them natural laboratories for studying how place shapes eating behavior. Thai society is undergoing rapid changes in food consumption patterns, driven by urbanization, the proliferation of convenience foods, and shifting work and study schedules. At the same time, non-communicable diseases linked to diet have become a central public health concern in Thailand and across the region. Understanding how eating behavior is socially patterned within a university therefore has implications that extend well beyond campus boundaries, because the habits formed or reinforced during university years often persist into later adulthood.

Among the study’s most consistent findings is the role of age. Older participants showed higher awareness of healthy eating, suggesting that nutritional consciousness accumulates over the life course rather than arriving fully formed in youth. This pattern challenges the assumption that younger generations, often portrayed as more health-literate, are inherently better positioned to make sound dietary choices. In the Thai university context at least, the data indicate that younger members of the population are more likely to occupy the less healthful subgroups, while awareness of healthy eating rises as age increases. For health promotion planners, this implies that interventions targeting younger students should not assume baseline motivation but must instead build it.

Income adequacy emerged as another significant determinant of food choices. The finding underscores a material dimension of healthy eating that is easy to overlook in behavior-focused campaigns: the ability to afford nutritious food shapes what people actually eat, independent of knowledge or intention. When budgets are constrained, cheaper HFSS options become more attractive, and the study’s regression results indicate that perceived income adequacy significantly shifts the odds of membership in the healthier eating classes. This aligns with a broader literature on food insecurity and diet quality, and it suggests that university policies on food pricing, subsidized healthy options, and campus food environments may be as important as education in shifting population-level eating patterns.

Perhaps the most actionable findings concern knowledge and behavior. Knowledge of food labels was identified as a key determinant of healthy eating awareness, as was body satisfaction. Participation in health-promoting activities and regular exercise also predicted membership in the more healthful subgroups. Taken together, these factors sketch a coherent picture: people who engage actively with their health, whether by reading labels, exercising, or joining wellness programs, are also the people who eat consciously and in balance. The relationship is likely reciprocal, with healthy behaviors reinforcing one another, but the practical implication is clear. Programs that build food label literacy and create opportunities for physical activity and health participation may generate spillover effects into dietary behavior.

The study was conducted by Phoobade Wanitchanon, Saichol Panyachit, Cholvit Jearajit, Pychaniphat Wichaino, and Poramin Tangopasvilaisakul, all of the Department of Sociology in the Faculty of Social Sciences at Srinakharinwirot University, and was supported by the Thai Health Promotion Foundation. Ethical approval was obtained from the university’s Human Research Ethics Committee, and the research was conducted in accordance with the principles of the Declaration of Helsinki. The authors note that the work responds to a gap in research on recent eating behaviors in Thai universities, where the intersection of student life, employment, and Thailand’s evolving food culture had not previously been examined with this class-based statistical lens.

Methodologically, the paper demonstrates the value of person-centered approaches in nutrition and social science research. Variable-centered methods, such as ordinary regression on averaged outcomes, can obscure heterogeneity by estimating effects across a blended population. Latent class analysis instead treats the population as a mixture of types, which is often closer to how health promoters encounter the real world: not as an average eater, but as distinct audiences with different motivations, constraints, and knowledge. The four-class solution identified here, validated through information criteria and followed by logistic regression on class membership, offers a template that other institutions in Thailand and the wider ASEAN region could adapt, particularly within networks such as the ASEAN University Network-Health Promotion Network.

The broader significance of the work lies in its framing of healthy eating as a social pattern rather than a purely individual choice. By showing that class membership is predicted by age, income adequacy, food label knowledge, body satisfaction, exercise, and participation in health-promoting activities, the study situates diet within a web of social and lifestyle factors that institutions can influence. Universities, as dense environments where diverse populations eat, work, and study together, are well positioned to act on these findings, whether through campus food policy, nutrition education embedded in curricula, or structured wellness programming. As food consumption patterns in Thai society continue to change rapidly, research of this kind provides the evidence base needed to design interventions that meet distinct groups of eaters where they actually are, rather than where an average model assumes them to be.

Subject of Research: Social patterning of eating behavior among students and personnel in Thai universities identified through latent class analysis.

Article Title: Social patterning of eating behavior in a Thai university population using latent class analysis

Article References: Wanitchanon, P., Panyachit, S., Jearajit, C., Wichaino, P., & Tangopasvilaisakul, P. (2026). Social patterning of eating behavior in a Thai university population using latent class analysis. Discover Social Science and Health. https://doi.org/10.1007/s44155-026-00476-6

Image Credits: AI Generated

DOI: 10.1007/s44155-026-00476-6

Keywords: eating behavior, latent class analysis, Thai university, healthy eating, food practices, non-communicable diseases, food labels, health promotion, sociodemographic factors, Thailand, university population, dietary patterns