**Background:** Obesity rates have nearly tripled since the 1970s, with 1.9 billion adults overweight and 650 million obese in 2016. Dining out is a context where unhealthy eating is particularly common—70% of meals in US fast-food restaurants and 50% in full-service restaurants are of poor nutritional quality. The unhealthy-tasty intuition posits that consumers associate unhealthy food with better taste, leading to lower consumption of healthy options. However, evidence on this intuition is inconclusive, with some studies finding no correlation or even a positive correlation between healthiness and taste expectations. This study examines how health claims and sensory claims on restaurant menus influence purchase intentions for healthy desserts, and whether health inferences and taste expectations mediate this relationship.
**Methods:** An online experiment with 273 participants randomly allocated to three conditions (no claim, health claim, sensory claim) was conducted on December 16, 2022. Participants were recruited via clickworker.com, were English-speaking UK residents aged 18-65, and completed the survey on desktop/notebook. After excluding 88 participants who failed attention checks and 48 with dietary restrictions relevant to dessert choice, the final sample comprised 137 participants (45% female, mean age 38.45 years, SD=10.10). Fictitious menus with four desserts were used; the target dish (berry yogurt crème with chocolate flakes) was manipulated with either a health claim ("low sugar" and "light") or sensory claim ("sweet" and "crunchy"). Measures included health inferences (3 items, α=0.90), taste expectations (2 items, r=0.84), purchase intention (3 items, α=0.94), and healthy eating habits (6 items, α=0.77). A serial multi-categorical mediation model (PROCESS model 6, 5,000 bootstrap samples, 95% CI) tested the hypotheses with healthy eating habits, age, and gender as covariates.
**Key Results:** The manipulation was successful: health claim condition participants agreed health information was used more (M_health=3.71 vs. M_control=2.17 vs. M_sensory=1.98, F(2,134)=10.22, p<0.01), and sensory claim participants agreed taste/consistency information was used more (M_sensory=4.74 vs. M_health=4.03 vs. M_control=3.26, F(2,134)=9.29, p<0.01). No significant differences in healthy eating habits across conditions (F(2,134)=0.82, p=0.44). Health claims significantly reduced taste expectations (a3=-0.65, p=0.04) and taste expectations increased purchase intention (b2=0.83, p<0.01), confirming a negative indirect effect on purchase intention via taste expectations (a3b2=-0.53, 95% CI[-1.04, -0.05]). Health claims significantly increased health inferences (a1=0.93, p=0.01), which increased purchase intention (b1=0.25, p<0.01), yielding a positive indirect effect (a1b1=0.23, 95% CI[0.06, 0.49]). Health inferences increased taste expectations (d21=0.31, p<0.01), and the serial indirect effect (health claim → health inferences → taste expectations → purchase intention) was significant (a1d21b2=0.24, 95% CI[0.06, 0.47]). Sensory claims did not significantly increase taste expectations (a4=0.25, p=0.41), and the indirect effect via taste expectations was not significant (a4b2=0.21, 95% CI[-0.32, 0.75]). The total effect of health claims on purchase intention was negative and significant (-0.80, p=0.04), with a significant negative direct effect (c'1=-0.74, p<0.01).
**Clinical Implications:** These findings challenge the effectiveness of health claims on restaurant menus for promoting healthy food choices, as the negative impact on taste expectations outweighs the positive effect of health inferences. However, the positive correlation between health inferences and taste expectations suggests that if consumers deliberately process health information, it may attenuate the negative effect. Sensory claims ("sweet," "crunchy") did not improve taste expectations, possibly due to ceiling effects or the use of specific taste descriptors rather than more appealing hedonic or indulgent labels. Policymakers and restaurant managers should consider alternative strategies—such as appealing food descriptors or menu composition changes—rather than relying solely on health claims to nudge healthier eating in dining-out contexts.