In recent years, the increase in lifestyle-related diseases has become a significant global health issue, requiring effective methods to improve daily life. This study proposes a route recommendation method to reduce ‘Food Approach Bias’ (e.g. food desire) during daily activities such as jogging or walking for health care. Food approach bias refers to the cognitive and behavioral bias in which individuals either “choose” or “avoid” food. According to the route recommendation for health care, it is unrealistic to avoid ubiquitous food shops, cafes, or restaurants. To solve this problem, our proposed method recommends shops on the route, such as cafes or restaurants, based on the user’s taste (specifically those with low calories), based on the users’ obesity levels and their logs’ histories of shop choices on the routes. The recommendation also aligns with three types of activities: normal walking, brisk walking, or jogging. This paper describes this recommendation approach, focusing on mitigating the Food Approach Bias relative to obesity levels, and evaluates the impact of user desire for food after the activity by recommending four types of shops on the routes, such as user tastes with high-calorie or low-calorie shops and opposite tastes with high-calorie or low-calorie shops.