Some individuals seek support around loneliness on social media forums. In this work, we aim to determine differences in the use of language by users-in different age groups and genders (female, male), who publish posts on Twitter expressing loneliness. We hypothesize that these differences in the use of language will reflect how these users express themselves and some of their support needs. Interventions may vary depending on the age and gender of an individual, hence, in order to identify high-risk individuals who express loneliness on Twitter and provide appropriate interventions for these users, it is important to understand the variations in language use by users who belong to different age groups and genders and post about loneliness on Twitter. We discuss the findings from this work and how they can help guide the design of online loneliness interventions.