How to deal with individual health frailty? The common answer is to mobilise clinic informations including her/his physical medical information as well as her/his mental medical information. Then, we have to consider simultaneously i) what can be observed at the same age in a subsample of individual with the same characteristics; ii) what can be observed in terms of trend over the time. In other words, the individual health status remains on her/his health curve. If we observe any change, the question is then how should we interpret this gap, from which value of the standard deviation should we consider a frailty prediction? We know today that the clinical determinants are just one dimension of the prediction of an individual's frailty. Consider social frailty is one of the other components. This project is aimed to study how the interaction of these two dimensions could be a good predictor of the individual frailty.