Introduction
Methods
Search and selection
Study selection
Data extraction and validity
Analysis
Role of the funding source
Results
Study characteristics
Author | Correlated NPA | No. | Mean age (SD) | % male | Years of education (SD) | Impaired cognition | substance abuse | Co-medication | % on cART | Type of cART | Duration of cART | Mean current CD4 (IQR) | Mean nadir CD4 (IQR) | Duration infection | No sign. difference on | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Caldwell (2014) | No | HIV+ | 34 | 46.1 (8.5) | 54 | 12.6 (1.8) | NR | 0 | NR | 79 | NR | NR | 550 | 201 | 7.6 years | Age sex |
HIV− | 28 | 44.9 (12.7) | 65 | 14.0 (3.4) | NR | 0 | NR | |||||||||
Thomas (2013) | Yes | HIV+ | 52 | 41 (14) | 90 | 14 (2) | 23 % impairment | 25 % | NR | 44 | NR | NR | 377 (291–616) | 260 (116–386) | NR | Age |
HIV− | 52 | 44 (14) | 51 | 15 (3) | NR | NA | NR | |||||||||
Ances (2010a) | Yes | HIV+ | 6 | 30 (7) | 100 | 15 (2) | GDS 0.34 | 16 % | NR | 83 | NR | NR | 757 (424–900) | 588 (438–750) | NR | Age sex education |
HIV− | 10 | 30 (6) | 60 | 18 (3) | NR | 0 | NR | |||||||||
Ernst (2009) | Yes | HIV+ | 31 | 49.6 (8.4) | 97 | 15.5 (2.2) | NR | 0 | No neuro-impairing medication | 100 | NR | NR | 415 (40.4) | 152 (24) | NR | Age sex education hematocrit |
HIV− | 32 | 46.9 (13) | 88 | 15.5 (2.3) | NR | 0 | No neuro-impairing medication | |||||||||
Melrose (2008) | Yes | HIV+ | 11 | 40.8 (7.1) | 100 | 16.3 (1.5) | 9 % mild impairment | 9 % alcohol | 36.4 % anti-depressant | 91 | NR | 9.9 (5.4) years | 694.2 (197) | NR | 9.9 (5.4) years | Age sex education |
HIV− | 11 | 40.9 (8.7) | 100 | 16.9 (1.8) | 0 | 9 % alcohol | 0 | |||||||||
Chang (2013) | Yes | HIV+ | 66 | 47.1 (8.6) | 100 | 14.6 (2.3) | 43.90 % HAND | 0 | NR | NR | NR | NR | 401.7 | 158.4 | 144.22 months | Age sex education hematocrit |
HIV− | 56 | 45.7 (12.7) | 100 | 14.8 (2.2) | 0 | 0 | NR | |||||||||
Ances (2008) | Yes | HIV+ | 24 | 45.5 (6.9) | 71 | 14.5 (3.5) | GDS 0.95 of 7 MND en 8 HAD | 0 | NR | 100 | 1* | 20 months | 368 | NR | >1 year | Age sex education |
HIV− | 10 | 46 (12.6) | 60 | 14 (3.2) | GDS 0.3 | 0 | NR | |||||||||
Chang (2008) | Yes | HIV+ | 24 | 40.15 (8.1) | 100 | 13.7 (2.6) | GCD art+:6.2(0.2) art−:6.0(0.9) | 0 | No neuro-impairing medication | 50 | 2* | 134.8 (21.7) months | 467.5 | 218.5 | 129.9 months | Age sex education hematocrit |
HIV− | 18 | 39.82 (12.3) | 100 | 13.9 (2.1) | GCD 2.7(0.5) | 0 | No neuro-impairing medication | |||||||||
Juengst (2007) | Yes | HIV+ | 31 | 47.7 (15.7) | 90 | 14.5 (2.7) | 35 % MNCD or HAD | NR | NR | NR | NR | NR | 397.41 | NR | NR | None |
HIV− | 16 | 42.3 (11.8) | 75 | 14.0 (3.0) | NR | NR | NR | |||||||||
Ernst (2002) | Yes | HIV+ | 10 | 36.3 (7.9) | 100 | 14.8 (2.0) | 0 | 20 % smoking | NR | 90 | 3* | NR | 375 (187) | 241 (145) | NR | Age sex education |
HIV- | 10 | 36.1 (6.8) | 100 | 15.6 (2.6) | 0 | 20 % smoking | NR | |||||||||
Chang (2001) | No | HIV+ | 11 | 41 (4.8) | 100 | 14 (2.1) | 36 % MCMD 27 % mild HIV-dementia | 0 | NR | 91 | 4* | NR | 329 (197) | 170 (126) | NR | Age sex education |
HIV− | 11 | 38 (4.8) | 100 | 16.4 (3.3) | 0 | 0 | NR | |||||||||
Maki et al. (2009) | Yes | HIV+ | 7 | 41.1 | 0 | 11.7 | NR | 29 | NR | 43 | NR | NR | NR | NR | NR | Age sex education |
HIV− | 4 | 42.8 | 0 | 12.3 | NR | 50 | NR | |||||||||
Chang (2004) | Yes | HIV+ | 18 | 38.2 (7) | 78 | 14.2 (1.6) | 55 % MCMD 33 % mild dementia | 0 | No chronic co-medication | 83.3 | PI-regime | 27.7 weeks | 287 (36) | 123 (37) | 91 (15) months | Age sex education hematocrit |
HIV− | 18 | 38.0 (8.8) | 78 | 14.5 (1.9) | 0 | 0 | No chronic co-medication | |||||||||
Castelo (2006) | Yes | HIV+ | 14 | 39 (9.1) | 100 | 14.8 (2.0) | 0 | 0 | 5* | 71 | NR | NR | 690 (370) data from 10 out of 14 patients | NR | NR | Age sex education |
HIV− | 14 | 40 (10.4) | 100 | 16 (1.9) | 0 | 0 | 5* | |||||||||
Ances (2010b) | No | HIV+ | 26 | 39 | 77 | 16 | NR | 0 | NR | 60 | NR | At least 3 months | 486 | 278 | NR | Age sex education |
HIV− | 25 | 41 | 56 | 15 | NR | 0 | NR | |||||||||
Schweinsburg (2012) | No | HIV+ | 11 | 41.8 (6.1) | 82 | 13.9 (2.4) | NR | 0 | NR | 91 | NR | NR | NR | 140 (29–300) | NR | Age sex education |
HIV− | 13 | 42.5 (14.5) | 77 | 14.9 (2.1) | NR | 0 | NR | |||||||||
du Plessis (2015) | Yes | HIV+ | 18 | 32 (4.6) | 11 | 11 (10–12)6*
| GDS 0.21 | 0 | NR | 0 | NA | NA | 433 (199) | NR | NR | Age sex education |
HIV− | 16 | 28 (5.2) | 6 | 12 (11–12)6*
| GDS 0.17 | 0 | NR | |||||||||
Ipser et al. (2015) | Yes | HIV+ | 15 | 40.6 (14.5) | 80 | 13.8 (2.2) | GDS 0.32 (0.28) | Alcohol 13.3 % | NR | 86 | NR | NR | 548 (255.29) | 310 (193) | 80.4 months | Age sex education |
HIV− | 15 | 39.7 (12.6) | 87 | 13.7 (1.3) | GDS0.29 (0.26) | Alcohol 46.7 % | NR | |||||||||
Ortega et al. (2015) | Yes | HIV+ | 131 | 39.0 | 72 | 13.1 |
Z-score −0.39 | NR | NR | 63 | NR | NR | 551 | 281 | NR | Sex education |
HIV− | 45 | 31.7 (10.9) | 58 | 13.4 (2.7) |
Z-score 0.14 | NR | NR |
Critical appraisal
Impact of HIV on BOLD characteristics
Study | HIV/SN | Region | Task | Software | Threshold | Correction M.C. | Results |
---|---|---|---|---|---|---|---|
Ances et al. (2011) | 6/10 | Visual cortex | Checker board | NR |
p = 0·05 | Mask used | HIV + showed reduction in mean functional BOLD changes over time and greater inter-subject variance in BOLD measures |
Ances et al. (2008) | 24/10 | Motor | Checker board + squeezing | Voxbo | NA (amplitude BOLD signal) | NA | No significant difference in BOLD amplitude between HIV + and − |
Juengst et al. (2007) | 31/16 | HRF | Finger tapping | NR | NA (HRF) | NA | No significant difference in mean peak values between HIV + and HIV− |
Ances et al. (2010a) | 26/25 | Visual cortex | Checkerboard | AFNI |
p = 0·05 | Yes, not specified | HIV+ reduced functional changes in BOLD signal |
Whole-brain and regions of interest analyses
Study | HIV/SN | Network | Task | Software | Statistical thresholding | Result |
---|---|---|---|---|---|---|
Caldwell et al. (2014) | 34/28 | Working memory | Sequential letter task | FEAT | FWE corrected at voxel level p < 0.05 later relaxed (not specified) | HIV+ greater activation on the simpler attention task but less activation on the working memory task |
Ernst et al. (2009) | 31/32 | Attention | Tracking balls | SPM2 | FWE corrected at voxel level p < 0.05 | HIV+ more activation in right prefrontal region only with the most difficult task |
Melrose et al. (2008) | 11/11 | Semantic event sequencing | Picture sequencing task + object discrimination control | SPM2 | Voxel threshold 0.001 uncorrected, small volume correction | HIV+ less signal change in frontal regions and left caudate and more signal changes in postcentral/supramarginal gyrus |
Functional connectivity: dysfunction within the basal ganglia and prefrontal cortex and within interactions between these regions | ||||||
Chang et al. (2013) | 66/56 | Attention | Tracking balls | SPM8 | FWE corrected at cluster level p < 0.05 | HIV+ has load-dependent decreased activation in right temporal region, while HIV− showed load-dependent increase |
Chang et al. (2008) | 24/10 | Attention | Tracking balls | SPM2 | FWE corrected at cluster level p < 0.05 used various thresholds | HIV+ has greater load-dependent activation in right frontal and cingulate regions |
Ernst et al. (2002) | 10/10 | Working memory | Sequential letter task | SPM99b | Voxel threshold 0.001 uncorrected | HIV+ has more BOLD activation in the lateral prefrontal cortex on all tasks |
Chang et al. (2001) | 11/11 | Working memory | Sequential letter + number task | SPM99b | Voxel threshold 0.001 uncorrected | HIV+ has greater activation in parietal regions and frontal lobes (lateral prefrontal cortex and supplementary motor area) |
Maki et al. (2009) | 7/4 | Memory | Encoding task, recognition task | SPM2 | Cluster corrected (min size >30) uncorrected threshold p < 0.05 | Encoding: HIV− more activation in hippocampal and temporal/frontal cortical structures. Recognition: HIV+ more in left superior temporal gyrus, hippocampus, and right insular cortex |
Chang et al. (2004) | 18/18 | Attention | Tracking balls | SPM99b | Cluster corrected for M.C. (not specified) | HIV+ decreased activation in the normal visual attention network and increased activation in adjacent/contralateral structures |
Castelo et al. (2006) | 14/14 | Memory | Encoding + recognition task | SPM99b | NR | Encoding: no difference. Recognition: HIV+ less activity in right posterior hippocampus, right inferior frontal gyrus, and left lingual gyrus and more activity in lateral frontal and posterior parietal regions |
Schweinsburg et al. (2012) | 11/13 | Fronto-striatal | Mental rotation task | AFNI | Cluster corrected multiple thresholds/cluster size | HIV+ had increased activation in areas of the PPC-striato-frontal pathway and in left insular and right occipital cortex and less activation in the anterior cingulate |
Plessis et al. (2015) | 18/16 | Ventral-striatal | Reward task | SPM8 | FWE corrected at cluster level p = 0.05 | No between group differences |
Study | HIV/SN | Network | Task | Software | Corrected for MC? | Pre-specified ROI? | Results |
---|---|---|---|---|---|---|---|
Thomas et al. (2013) | 52/52 | Functional connectivity 5 domains | Resting state | NR | FDR corrected p < 0.05 | Yes | HIV+ had less intra- and internetwork correlations in several functional brain networks |
Chang et al. (2008) | 24/18 | Visual attention | Tracking balls | SPM2 | Uncorrected p = 0.05 | No | HIV+ load-dependent increase in frontal regions when HIV− has load-dependent decrease |
Maki et al. (2009) | 7/4 | Episodic encoding | Encoding task, recognition task | SPM2 | Cluster corrected | Yes | HIV+ decreased hippocampal activity during encoding and increased hippocampal activation during recognition |
Castelo et al. (2006) | 14/14 | Episodic encoding | Encoding task, recognition task | SPM99b | Not reported | Both | HIV+ had attenuated activation of brain regions known to support episodic encoding (right posterior hippocampus, left and right lingual gyrus, right inferior frontal gyrus) and recruited additional cortical regions |
Hippocampal activation; no | Yes | HIV+ less activation in bilateral hippocampus | |||||
Plessis et al. (2015) | 18/16 | Ventral-striatal reward | Reward task | SPM8 | no | Yes | HIV+ decrease in activation in ventral striatum for anticipating neutral and rewarding cues |
Ortega et al. (2015) | 132/49 | Functional connectivity 4 domains | Resting state | FS-FAST | FDR corrected p = <0.05 | Yes | HIV+ had lower cortico-striatal functional connectivity. HIV+ cART+ had higher connectivity than HIV+ cART− |
Ipser et al. (2015) | 15/15 | Functional connectivity 3 domains | Resting state | AFNI | Not reported | Yes | HIV+ had reductions in connectivity in fronto-striatal regions. |
Effect of cART
Effect of aging
Study | HIV/SN | Network | Task | Software | WB/ROI | Statistical inference | Results |
---|---|---|---|---|---|---|---|
Thomas et al. (2013) | 52/52 | Functional connectivity 5 domains | Resting state functional connectivity | NR | ROI | FDR corrected threshold of 0.05 | Aging causes decrease in intranetwork correlations in DMN and SAL and internetwork correlations between DMN-SAL. No interaction between HIV and aging |
Ernst et al. (2009) | 31/32 | Visual attention | Tracking balls | SPM2 | ROI | FWE corrected at voxel level p < 0.05 | After 1 year, HIV+ more BOLD signal in right prefrontal and posterior parietal cortices and cerebellum bilaterally. HIV− less BOLD signal after 1 year |
Chang et al. (2013) | 66/56 | Visual attention | Tracking balls | SPM8 | WB | FWE corrected at cluster p < 0.05 | HIV+ had greater age-related increases in brain activation in right parietal, cingulate and paracentral regions, cerebellar vermis, left frontal, temporal and occipital regions |
Juengst et al. (2007) | 31/16 | HRF | Finger tapping | NR | WB | NA (HRF) | No effect or interaction with HIV status for age in mean BOLD peak value |
Ances et al. (2010b) | 26/25 | Visual cortex | Checkerboard | AFNI | VOI |
p = 0.05 corrected for M.C. (not specified) | HIV and increasing age independently caused decreases in functional BOLD signal, no interaction |
Ipser et al. (2015) | 15/15 | Functional connectivity 3 domains | Resting state | AFNI | ROI | Not reported | Reduction in connectivity in individuals over 50 years, no interaction between age and HIV |