Changes in network composition among older people living in inner London and Essex

Ann Bowling, Morag Farquhar, Emily Grundy

Research output: Contribution to journalArticlepeer-review

15 Citations (Scopus)


This article examines changes over 212–3 years in social network composition among three samples of elderly people, and reports on a considerable amount of change in network size and structure. The samples were based in an urban section of inner east London (Hackney) (two samples of people aged 65–84 and 85+ at baseline) and in a semi-rural district of Essex (Braintree) (one sample aged 65–84 at baseline). Most people in all three samples had a social network, although 14% of the 85+ group had no one in their networks by follow-up. Loss of relatives, friends and confidants was greatest, as would be expected, in the 85+ sample. This older sample not only had weaker networks, and suffered the greatest losses to their networks, but they were also far less likely to be engaged in social activities. The younger Braintree sample members had the greatest degree of involvement in social activities, and they also had stronger network structures. Although most members of all three samples could name a main helper, sizeable proportions (21 and 16% of the younger and older Hackney samples, respectively) reported that they had no helper by follow-up, in comparison with just 5% of Braintree sample members. Both the Hackney samples were more likely than the Braintree sample to report wanting more help. The longitudinal analyses showed evidence of gains, as well as stability and loss, among sizeable proportions of sample members. Even among the 85+ sample, of those with no friends at baseline, 34% reported having friends by follow-up. Gains in friendships were also reported by the two younger sample members. However, gains should not be emphasized at the expense of losses, which were considerable. Network losses were due mainly to loss of relatives, rather than friends. Multivariate analysis confirmed that the greater part of the change in network size was accounted for by changes in numbers of relatives. Increases in network size over the time period analysed were associated with increases in need for help—thus size partly expanded in response to need for help. In the older 85+ sample, increases in network size were associated with poorer functional ability over time. The implication of the research for health and social policy is that networks do respond to needs, but the greater loss of relatives from the network over time can have important consequences given that relatives, not friends, provide the practical help.
Original languageEnglish
Pages (from-to)149-166
Number of pages18
JournalHealth & Place
Issue number3
Publication statusPublished - Sep 1995


  • old age
  • social support
  • social network
  • life satisfaction

Cite this