The Question That Started It All
Every day, millions of Sri Lankans scroll past health messages on Facebook posted by government health institutions. These posts warn about sugar, promote physical activity, and encourage healthy diets. They look professional, they reach large audiences, and they are produced with good intentions. But are they actually designed to change behaviour?
That was the question that set this study in motion. Not whether the posts were visually appealing, and not whether they reached a large number of people, but whether they were built on the kind of theoretical foundations that decades of health communication research tells us are necessary to actually shift what people think, feel, and do about their health.
Two Frameworks, One Big Gap
To answer this question, we needed a way to evaluate Facebook health content that went beyond counting likes. So we developed two evaluation frameworks: one based on the Health Belief Model (HBM), which examines whether content addresses perceived susceptibility, severity, benefits, barriers, cues to action, and self-efficacy; and another combining Lasswell's 5W model with the Social Marketing Mix, which evaluates how strategically the message is structured, targeted, and delivered.
We applied these frameworks to 15 purposively selected Facebook posts from three official Ministry of Health pages in Sri Lanka: the NCD Unit, the Health Promotion Bureau, and the Nutrition Division. The posts covered physical inactivity and unhealthy diet, the two leading modifiable NCD risk factors in the country.
The finding was clear and a little uncomfortable. The posts were generally strong strategically: they targeted audiences, used appropriate visuals, and structured their messages around clear health topics. But they were significantly weaker in their application of health communication theory. Most posts failed to engage the psychological constructs of the Health Belief Model in any meaningful way, particularly perceived barriers and self-efficacy, which are the dimensions most directly linked to behaviour change.
The Animated Video Exception
There was one content type that stood out: animated videos. These consistently achieved the highest theory scores, ranging from 88.8% to 94.4%, and also performed better strategically. The highest-scoring post of all was a paid animated video on unhealthy diet, which scored 94.4% on the theory evaluation framework.
But here is the catch. That same post reached 132,000 people and generated only 93 engagements. A still image posted organically on the same page reached 51,776 people and generated 8,441 engagements. This tells an important story: theoretically sound content does not automatically generate engagement. And high engagement does not mean the content is theoretically grounded.
This challenges a common assumption in digital health communication: that engagement is a proxy for effectiveness. It is not. A post can go viral for reasons that have nothing to do with whether it will change the dietary habits of the people who saw it.
What Was Missing: Physical Inactivity
A second finding that surprised us was the near-absence of physical activity content. Of the 137 posts we screened across four years and three Facebook pages, only a small proportion addressed physical inactivity, despite it being one of the most significant modifiable risk factors for NCDs in Sri Lanka. Unhealthy diet dominated the content landscape almost entirely.
This likely reflects a practical communication challenge: diet-related messages translate easily into visually appealing formats. A plate of colourful food, a recipe, a label guide. Physical activity is harder to communicate compellingly in a static image or short video without production investment. The result is a structural imbalance in the NCD prevention communication landscape that mirrors trends observed in the UK and other countries, and that deserves deliberate policy attention.
A Study Born Out of a PhD
This research sits within my doctoral work at the University of Kelaniya, Sri Lanka, supported by the NIHR Global Health Research Unit on Health Communication in South Asia, in partnership with Imperial College London. The study emerged from a straightforward observation: Sri Lanka has a sophisticated public health infrastructure, a significant NCD burden, and a government that is genuinely committed to prevention. Yet the digital communication produced by that infrastructure had never been systematically evaluated against the theoretical standards the field has developed over decades.
The frameworks we built in this study were developed from an extensive scoping review of health communication theories and strategies, the full methodology of which is reported in a companion manuscript currently under review at BMC Public Health. Together, these two papers form the core of my doctoral research on how theoretical grounding and strategic design can make digital health communication more effective for NCD prevention.
Health communicators, content creators, and policymakers in Sri Lanka and across the South Asian region now have a practical, evidence-based set of frameworks they can use to evaluate and improve their own digital health content. The frameworks are not complex, and they are not proprietary. They are built on theories that have been developed and validated over decades, applied to a context where they have rarely been used systematically.
The message from this study is not that Sri Lanka's health communication is failing. It is that it has the foundations of something genuinely effective, and that closing the gap between strategic strength and theoretical depth is both achievable and important. With the right training for content creators, the right policy mandate for theory-informed communication, and the right evaluation mechanisms, social media can become a far more powerful tool for NCD prevention than it currently is.
The posts are already reaching people. Now we need them to change behaviour.