Inflammation alone didn't clearly dent memory. With insomnia, it did.

Randomised controlled trial (RCT). 84 people. one lab day; memory tests 2 hours after the injection (single time point).

A trial. Chance decided who did what.

Who: Non-depressed adults aged 60-80 in good general health in Los Angeles (mean age 66.0, 50% women, 79.8% White); 26 with DSM-5 insomnia disorder, 58 without insomnia. A subsample of the parent SHARE-D trial (NCT03256760) who consented to the spatial tests. Excluded: major medical/psychiatric illness, depression in the past year, sleep apnea, BMI > 35, smoking, psychotropic drugs

• Insomnia: maze memoryworsevs good sleepers
• Inflammation shot onlynoneno clear change
• Insomnia + shot: mapworsevs dummy shot

More to know, from our read of the paper:

• Insomnia was not randomized: the insomnia-vs-good-sleeper differences are a group comparison (linked, not proven). Only the shot (endotoxin vs placebo) was randomized
• Secondary outcomes of a trial whose primary outcome was depressed mood (reported separately, Irwin et al. 2025). The authors call the results hypothesis-generating
• Small groups: 12 insomnia + endotoxin, 14 insomnia + placebo, 27 control + endotoxin, 31 control + placebo. The key interaction is P < 0.05 with post hoc tests
• Spatial memory was measured two ways; insomnia was worse only on platform-site crossings (P < 0.01), not on % time in the target quadrant (P = 0.08). Say 'worse on the maze memory test', never 'worse on every memory test'
• The shot alone had no clear effect on any outcome (no main effect of condition), including in good sleepers. Inflammation levels (IL-6, TNF-alpha) did not track the memory scores, so the mechanism is unproven

84 healthy adults aged 60 to 80 took part, 26 of them with insomnia. In a randomized trial, each got either a shot that briefly sparks inflammation (low-dose endotoxin) or a dummy shot, then did a virtual maze and rebuilt the room from memory two hours later. People with insomnia did worse on the maze memory test and on rebuilding the room than good sleepers. The inflammation shot alone made no clear difference, but with insomnia it went with a worse room score. Insomnia itself wasn't randomized, so that part is linked, not proven, and it was one test on one day.

We read the abstract and the full text of this paper and checked every number on this page against it (16 facts, checked on 2026-10-29). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.