Clinical studies suggest people with Alzheimer's disease (AD) have altered pain sensitivity. Experimental pain research is equivocal.\ud
Conduct a meta-analysis to investigate if people with AD have altered pain sensitivity compared to healthy controls (HCs).\ud
Three authors searched electronic databases from inception till November 2015 for experimental pain studies in AD vs. HCs. Outcome measures were pain threshold, tolerance, pain ratings, heart rate response to noxious stimuli and the Facial Action Coding System (FACS). Random effect meta-analysis calculating Hedges' g±95% confidence intervals (CI) was conducted.\ud
Thirteen studies were identified, including 256 people with AD (74.6 (±5.6) years, 59% females with a mean mini mental state examination (MMSE) score of 19.2) and 260 HCs. Meta-analysis demonstrated no significant difference in pain threshold (g=0.025, 95% CI -0.315-0.363, p=0.88, n AD=135, n HCs=157), pain tolerance (g=-0.363, 95% CI -2.035-1.309, p=0.67, n AD=41, n HCs=53) or pain intensity ratings (g=0.03, p=0.89, n AD=138, n HCs=135). Heart rate response to pain was less pronounced in AD but not significant (g=-0.746, p=0.11). People with AD (n=90) had significantly higher FACS scores versus HCs (n=109) (g=0.442, p=0.03) indicating increased pain. Meta-regression demonstrated that an increasing percentage of AD female participants moderated pain threshold (p=0.02) whilst MMSE scores did not (p=0.19).\ud
People with AD have a greater sensitivity to pain when validated observer ratings of facial expressions are used. Verbal response to painful stimuli, even under experimental conditions, may mean pain is not identified in people with AD. Clinically useful observer rated pain tools may be the most appropriate way to assess pain in AD