Alcohol consumption is associated with a wide variety of preventable health complications and is a major risk factor for all-cause mortality in the age group 15-47 years. To reduce dangerous drinking behavior, eHealth applications have shown promise. The most advanced such eHealth applications make use of underlying predictive models. However, existing mathematical models do not consider real-life situations, such as combined intake of meals and beverages, and also do not connect drinking to clinical markers, such asphosphatidylethanol(PEth). Herein, we present such a model. More specifically, we have developed a new sub-model for gastric emptying, which depends on all food and beverages consumed. The new model can accurately describe both training data and independent validation data, not used for training. The model can also be personalized using e.g. anthropometric data from a specific individual and can thus be used as a physiologically based digital twin. This twin is also able to connect short-term consumption of alcohol to the long-term dynamics of plasma PEth. We illustrate how this connection allows for a new way to determine patient alcohol consumption from PEth levels, which has the potential to improve upon traditionally used self-reporting forms. Finally, the new model is integrated into a new eHealth app, which also could help guide individual users or clinicians to help reduce dangerous drinking.