TY - JOUR T1 - RF37 Predictors of loneliness among older men and women in england JF - Journal of Epidemiology and Community Health JO - J Epidemiol Community Health SP - A60 LP - A60 DO - 10.1136/jech-2018-SSMabstracts.125 VL - 72 IS - Suppl 1 AU - H Yang AU - PA Bath Y1 - 2018/09/01 UR - http://jech.bmj.com/content/72/Suppl_1/A60.1.abstract N2 - Background Loneliness has been considered as a serious social and public health problem among older people. Understanding factors that affect loneliness among older adults is particularly important. However, Loneliness is a complex and multidimensional construct, which is associated with a wide variety of factors. A consideration of multiple dimensions is essential to gain a complete picture of an individual’s state of loneliness.Objective The objective of this study is to investigate the prospective impact of a wide range of factors on loneliness with a focus on: (a) which factors are closely related to the feelings of loneliness, and whether they perform consistently as people age, and (b) whether men and women differ in loneliness experiences.Methods Data used in the study were from a population-based sample of 3838 core members who participated in all the waves of the English Longitudinal Study of Aging (ELSA) (2002–2015). Binary logistic regression analyses were conducted to understand the relationship of loneliness and associated factors on two sets of experiments: (a) four different wave points with a 4-year measure span were selected to examine the effect of different groups of factors on loneliness prediction and potential prediction changes when ageing. (b) The predictive models were built on separate gender datasets to investigate whether the risk factors for men and women were different. Statistical software SPSS (version 23) was used in the analyses.Results Variables in the models included: gender, age, marital status, closeness to spouse, contact with children and friends, money shortage, health problem, and depression. The results on four wave points showed that widowhood, emotionally distant to spouse, and depression displayed consistently significant association with loneliness. But contact with friends, money shortage and health problem only showed significance on one or two waves. It suggested there might be changes in risk factors for loneliness as people age. The results on separate gender data confirmed that being widowhood or not close to partner led to a higher risk of being loneliness for women, while health problem and infrequent contact with friends increased the feelings of loneliness in men.Conclusions This study identified important risk factors influencing loneliness experience. The findings from the study showed some evidence that the impact of risk factors might change when ageing or in different gender. Knowledge about such difference will be helpful in the development of targeted interventions to fight against loneliness in later life. ER -