Population confinements are an unprecedented situation that entails very significant constraints on social interactions (talking to each other, working together, carrying out common activities, etc.). In France, the duration of the spring confinement (a little less than two months) is short in terms of interpersonal relationships, but these relationships were put to the test and may have been altered, withdrawn or, on the contrary, strengthened, while new links were being established. We can therefore ask ourselves to what extent, with the establishment of the crisis in the long term, relationships are not only momentarily tested, but are more durably affected. One of the objectives of the survey "Life in Confinement" launched in April 2020 by a group of sociologists (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3807975) was precisely to study the effects of the crisis on interpersonal relations. The first wave of the survey (April 14-May 10) included a series of questions about relationships, which allowed us to make an initial analysis of the strain that confinement has placed on this aspect of social life. We observed that the relationships of the young people were particularly affected in a context where outings and other social activities, which are particularly frequent in this age group, were prevented. More generally, we observed that people were preoccupied with maintaining or strengthening relationships with close friends (family, certain friends) and tended to sort out their other relationships, with a reinforcement of the "entre-soi" of the most qualified. Moreover, women seemed to be more involved than men in maintaining all kinds of relationships. Of the 16224 people who responded to our first questionnaire, 5434 had left an e-mail address and expressed an interest in continuing the survey. We sent them a second questionnaire between December 17 and January 26 and collected 3620 responses. This second questionnaire of course included questions about changes in personal relationships, based on the responses to the first questionnaire. The in-depth analysis of the data from the first questionnaire, reinforced by that of the data from the second, now allows us to propose a perspective on the effects of the crisis on personal relationships. In this paper, we start by presenting the data we have on personal relationships in the two questionnaires, as well as the characteristics of the samples. Then, we will show that age (coded in brackets allowing to capture approximately large life cycles) is the variable most correlated with those concerning variations in the loss or deterioration of interpersonal relationships, more than the level of education, the professional category, the gender or the configuration of confinement. An exploration by these broad age categories will show the differentiated effects of the crisis on interpersonal relationships.