P125 The relationship between mid-life pain and cognitive impairment: a prospective, population-based study

Abstract

Abstract Background/Aims Population-based studies suggest an association between pain in elderly adults and cognitive impairment. However, the temporal relationship between chronic pain and cognitive decline is uncertain: most studies are cross-sectional or have short follow-up periods. Our aim was to examine the association between two pain phenotypes - bodily pain and multi-joint pain - in middle-life and cognitive impairment in later-life in a prospective cohort study with 23 years’ follow-up. We also set out to examine the association with mortality. In contrast to much previous work, this study has available a comprehensive depth of information on health and pain characteristics over a long follow-up period. Methods Bodily pain and multi-joint pain at 3 timepoints between baseline and year 10 was measured in participants from the Chingford Women’s Study. Secondary analyses were performed on the two dimensions of bodily pain, pain intensity and interference. Multivariable logistic regression analysis was used to assess the association with cognition at year 23: the primary cognitive outcome was Mini-Mental State Examination (MMSE) score ≤28, with a secondary outcome of ≤ 26. Multivariable Cox regression analysis was used to evaluate the association between each pain exposure and all-cause mortality between years 10 and 23. Analyses were adjusted for education, socio-demographic factors, and modifiable risk factors for cognitive impairment including depression and analgesia use. Results 282 (28.1%) and 276 (27.5%) women were included in analyses for cognition and bodily pain and multi-joint pain respectively. Women with severe bodily pain showed a non-significant greater odds of MMSE ≤28 compared to those with no/mild bodily pain (OR 2.56, 95%CI 0.94-6.95, P = 0.07). Those with severe bodily pain had significantly greater odds of MMSE ≤26 (OR 7.31, 95%CI 2.11-25.36, P = 0.002). However, there was no association with multi-joint pain. Secondary analysis of the two dimensions of bodily pain demonstrated that pain interference displayed a stronger relationship with both MMSE thresholds compared to pain intensity. Women with severe bodily pain and repeated multi-joint pain each had higher crude all-cause mortality, but this was not statistically significant (log-rank test for trend, P = 0.11 & 0.068, respectively). Further adjustment left the association with both pain exposures and all-cause mortality entirely attenuated. Conclusion After 13 years’ follow-up, severe bodily pain but not multi-joint pain in mid-life was associated with subsequent cognitive impairment, determined by MMSE ≤26. Women with severe bodily pain and multi-joint pain had higher crude all-cause mortality rate, but this was not statistically significant after adjustment. Pain interference displayed a stronger association with cognitive impairment; this may reflect the impact of pain on functional status. This study represents the first time the consequences of mid-life pain on later life cognition were examined. Chronic pain, which is largely untreated, may thus represent a modifiable cause of cognitive impairment. Disclosure E. Kelleher: None. S. Hawley: None. A. Delmestri: None. N. Arden: None. I. Tracey: None. A. Soni: None.