Long-term care (LTC) delivered to elderly persons in need of assistance in activities of daily living is a topic of increasing importance. The financing of LTC, the needs for specialized infrastructure and the limited number of caregivers will pose a systemic threat in many developed countries. In this paper, we analyze the factors influencing the old-age care prevalence rates in Switzerland through a log-linear regression model. Based on a cross-sectional dataset covering the LTC needs from 1995 to 2014, we statistically support the effect of key drivers such as the age, the gender and the region of residence. We distinguish the prevalence by the mild, moderate and severe frailty levels and by care received either at home or in an institution. Our regression results evidence that prevalence rates exponentially increase with the age yielding significantly higher values for women. These effects are emphasized for moderate and severe dependence and for institutional care. Finally, we forecast the number of dependent persons until 2045. Our projections reveal an important increase in the future numbers. While we observe that the dependent population more than doubles over the 30-year horizon, we report significant cantonal differences. Our results are relevant to governments, practitioners and academics alike and help to better understand the factors affecting the demand of LTC and predicting future needs.