Abstract
The finding of a potential association between pain and cognitive decline is limited to a few cross-sectional studies with relatively samples. We therefore aimed to investigate whether the presence and severity of pain at baseline could predict a decline in cognitive function over four years of follow-up in the English Longitudinal Study of Ageing. At baseline, participants with no dementia who were "often troubled by pain" were considered to have pain. Pain severity was categorized as mild, moderate, or severe. Cognitive function was explored through verbal fluency (assessed by asking how many different animals the participants could name in 60 s), memory (sum of immediate and delayed verbal memory) and processing speed (number of target letters correctly identified on the letter cancellation task). Multivariable linear regression was used (exposure: pain; outcomes: cognitive change between follow-up and baseline, based on standardized residuals). Altogether, 6515 community-dwelling people with a mean age of 65 years (women = 57.3%) were included. Over a 4-year follow-up, after adjusting for 26 potential confounders, no association between pain (yes vs. no) and verbal fluency (beta = 0.02; 95%CI: -0.15 to 0.18), memory (0.05; 95%CI: -0.28 to 0.38), or processing speed (0.55; 95%CI: -18.4 to 2.93) at follow-up was found. Only severe pain was associated with greater decline in memory (-0.36; 95%CI: -0.68 to -0.04). In conclusion, in older people, pain was not associated with worsening in cognition, except for severe pain, which was marginally associated with worsening in memory tests. Further longitudinal studies are needed to confirm or refute our findings.