Effect of social support on memory ageing of middle-aged and older cancer survivors: a marginal structural modelling approach.

Background: While social support is associated with better cognitive health among cancer-free individuals, this relationship is understudied among cancer survivors. We investigated whether overall social support before and after a cancer diagnosis is related to post-diagnosis memory ageing, overall and by sex/gender.

Methods: Data were from 2044 cancer survivors in the US Health and Retirement Study (HRS; n=1395) and English Longitudinal Study of Ageing (ELSA; n=649) from 2006 to 2018. Incident cancer diagnoses and memory function (immediate and delayed word recall) were assessed biennially. Social support was assessed every 4 years in the HRS and biennially in ELSA. We established three time points relative to a cancer diagnosis: pre-diagnosis (the wave prior to cancer diagnosis), time 1 post-diagnosis (the first wave after a cancer diagnosis) and time 2 post-diagnosis (the second wave after a cancer diagnosis). Multivariable-adjusted marginal structural models incorporating inverse probability of treatment and attrition weights estimated the relationship between overall social support and memory function post-diagnosis.

Results: Prior to a cancer diagnosis, 45.1% of participants reported high social support. Cancer survivors reporting higher social support at time 2 had better memory function post-diagnosis than those with lower social support (0.14 SD units; 95% CI: 0.03 to 0.24) which was stronger among women (0.18 SD units; 95% CI: 0.02 to 0.34) than men (0.10 SD units; 95% CI: -0.03 to 0.24).

Conclusions: Social support may help promote memory function after a cancer diagnosis in mid-to-later life. Further studies with a larger sample size and differentiation of social support are warranted.