Table 3.

Selective evidence on the association between lipopolysaccharides and cognitive function.

Clinical Trials
Reference	Sample Characteristics at Baseline	Endotoxemia Assessment	Outcomes	Study Design	Results
Krabbe KS. et al., Brain Behav Immun, 2005 [59]	12 healthy subjects (mean age 26 y, all men)	Injection of LPS E. coli 0.2 ng/kg or placebo	Memory and learning (Word-list memory test) Working memory (Digit Span backward and Letter–Number Sequencing) Attention (Digit span forward and Digit symbol) Executive functions (Trail Making Test A and B) Assessment before and at 1.5, 6 and 24 h post-injection	Randomized double-blind crossover study	Negative correlation between IL-6 at 4.5 and 6 h post-injection and the word-list learning performance LPS injection did not affect performance on other cognitive tests
Kullmann JS et al., Soc Cogn Affect Neurosci, 2014 [64]	18 healthy subjects (mean age 26 y, all men)	Injection of LPS E. coli 0.4 ng/kg or placebo	Emotional/social processing (Reading the mind in the eye) Brain activation during the test Assessment at 2 h post-injection	Randomized double-blind crossover study	LPS injection did not affect the number of correct responses during the Reading the mind in the eye test  Responses in the fusiform gyrus, temporo-parietal junction, superior temporal gyrus and precuneus regions compared to placebo
Grigoleit JS. et al., Neurobiol Learn Mem, 2010 [61]	24 healthy subjects (mean age 25 y, all men)	Injection of LPS E. coli 0.4 ng/kg or placebo	Attention and executive functions (Color word stroop task (assessment before and 1.5 h post-injection)) Verbal, visual and delayed memory (Revised Wechsler Memory Scale (assessment at 3 h post-injection))	Randomized double-blind controlled trial	LPS injection did not affect performance on cognitive tests
Grigoleit JS. et al., PLoS One, 2011 [62]	18 healthy subjects in the low-dose group (LPS 0.4 ng/kg) and 16 subjects in the high-dose group (LPS 0.8 ng/kg), mean age 25 y	Injection of LPS E. coli 0.4 ng/kg or placebo, or LPS E. coli 0.8 ng/kg or placebo	Working memory (N-back task) Assessment at 1.75 and 3 h post-injection	Randomized double-blind crossover study	Reaction time in N-back task in the high-dose group compared to placebo LPS did not affect the accuracy in N-back task
Moieni M et al., Brain Behav Immun, 2015 [63]	109 healthy subjects (mean age 24 y, 40% men): 58 subjects in the LPS injection group and 51 in the placebo group	Injection of LPS E. coli 0.8 ng/kg or placebo	Emotional/social processing (Reading the mind in the eye) Assessment at 1 h 40 post-injection	Randomized double-blind controlled trial	Number of correct responses in the LPS group
Reichenberg A. et al., Arch Gen Psychiatry, 2001 [57]	20 healthy subjects (mean age 24 y, all men)	Injection of LPS Salmonella abortus equi 0.8 ng/kg or placebo	Declarative memory (Story recall, Figure recall and Word-list learning) Attention (Digit span forward, Digit symbol, Ruff 2 and 7 cancellation test, Simple reaction time test and Continuous performance test) Executive functions (Colored Trail Making Test A and B and Word fluency test) Assessment at 1–2, 3–4 and 9–10 h post-injection	Randomized double-blind crossover study	Performance on story recall, figure recall and word-list learning tests compared to placebo at 1–2, 3–4 and 9–10 h post-injection LPS injection did not affect performance on other cognitive tests
Cohen O. et al., J Mol Neurosci, 2003 [58]	10 healthy subjects (sub-sample of the Reichenberg’s study [57])	Injection of LPS Salmonella abortus equi 0.8 ng/kg or placebo	Declarative memory (Story recall) Working memory (Digit span backward) Attention (Digit span forward and Ruff 2 and 7 cancellation test) Assessment at 1–2, 3–4 and 9–10 h post-injection	Randomized double-blind crossover study	Performance on story recall compared to placebo at 1–2, 3–4 and 9–10 h post-injection  Performance digit span backward test compared to placebo at 1–2, 3–4 and 9–10 h post-injection LPS injection did not affect performance on other cognitive tests
Van den Boogaard M. et al., Crit Care, 2010 [60]	15 healthy subjects in the injection group (mean age 23 y, all men) and 10 healthy controls (mean age 25 y, all men)	Injection of LPS E. coli 2 ng/kg or no injection	Working memory (Digit span backward) Attention (Digit span forward, Color word stroop task and Paced Auditory Serial Addition Test (PASAT)) Psychomotor speed capacity and information processing ability (number of correct response on the Digit symbol test) Fine control motor (time required to finish the Grooved pegboard test) Assessment at 1–2, 3–4 and 9–10 h post-injection	Single-blind trial	Performance on PASAT test compared to control groupLPS injection did not affect performance on other cognitive tests
Observational studies
Reference	Sample characteristics at baseline	Endotoxemia assessment	Outcomes	Study design	Results
Lyons JL. et al., J Acquir Immune Defic Syndr, 2011 [66]	97 HIV-infected patients (mean age 47 y, 76% men)	Plasma LPS quantified by LAL	HIV-associated neurocognitive disorders (HAND) Global T test (motor skills, speed of information processing, attention (working memory), learning (memory encoding), memory (memory recall), language fluency, and executive function)	Cross-sectional	LPS levels did not differ according to the severity of HAND LPS levels did not differ in subjects with global impairment (Global T test < 40) compared to unimpaired subjects
Vassallo M. et al., J Neurovirol, 2013 [65]	40 HIV-infected patients with HIV-associated neurocognitive disorders (HAND) (median age 46 y, 78% men) and 139 HIV-infected patients without HAND (median age 44 y, 72% men)	Plasma LPS quantified by LAL	HIV-associated neurocognitive disorders (HAND) z-score (learning and recall episodic memory, attention/concentration, working memory, executive functions, language, visual agnosia and motor/psychomotor speed)	Cross-sectional	LPS levels was higher in the HAND group compared to no-HAND group LPS levels did not differ according to the severity of HAND
Jespersen S. et al., BMC Infect Dis, 2016 [67]	62 untreated HIV-infected patients without evidence of impaired cognitive function (mean age 39 y, 52% men)	Plasma and CSF LPS quantified by LAL	CSF neurofilament light chain protein (marker of CNS axonal damage) and CSF neopterin (marker of monocyte activation)	Cross-sectional	No association between plasma LPS and CSF neurofilament light chain protein or CSF neopterin LPS was not detectable in CSF

Abbreviations: AD Alzheimer’s Disease; LPS Lipopolysaccharide; LAL Limulus Amebocyte Lysate; HIV Human Immunodeficiency Viruses; HAND HIV-associated neurocognitive disorders; CSF Cerebrospinal Fluid; CNS Central Nervous System; NFL Neurofilament light chain protein.