Table 2.
Quality assessment of studies included in the meta-analysis
| Barnes et al. (2007b) | Du et al. (2004) | Fox et al. (2005) | Hashimoto et al. (2005) | Jack et al. (2004) | Jack et al. (2003) | Kaye et al. (2005) | Thompson et al. (2004) | Wang et al. (2003) | |
|---|---|---|---|---|---|---|---|---|---|
| Study type | Case series with controls | Case series with controls | Trial | Case series | Case series with controls | Trial | Case series with controls | Case series with controls | Case series with controls |
| AD Diagnostic criteria (imaging used in criteria) |
NINCDS–ADRDA | NINCDS–ADRDA (imaging used to exclude other neuropathol.) |
NINCDS–ADRDA (imaging supporting AD) |
NINCDS–ADRDA (MMSE >15) |
NINCDS–ADRDA/DSM IIIR (general committee consensus) |
Mild–moderate AD (>50 years old) |
NINCDS–ADRDA (mixture of community and clinical cohorts) |
NINCDS–ADRDA/DSM- IV |
DAT = CDR of 0.5 |
| AD exclusion criteria | Change in diagnosis at subsequent follow-up or post-mortem confirmation of a differing disease |
Other neuropathol., head trauma, alcoholism, psychiatric illness, epilepsy, hypertension, diabetes, major heart disease |
Significant other neurological diseases. Medications which affect cognition |
Other neuropathol./mental disorders/congnitive-affecting treatments/focal brain lesions seen on MRI |
Symptoms unrelated to AD, depression |
Non-AD disorders, major and/or unstable conditions such as seizure, PD or tumour |
Substance abuse, depression, Parkinson’s disease, depression blood test used to screen for exclusionary medical conditionsa |
Substance abuse, depression, psychiatric illness/head injury |
Other confounding neuropsychol. Genetic family history of AD |
| Specific white matter damage exclusion criteria |
N/R | N/R | N/R | N/R | Stroke | Substantial cerebro-vascular disease |
N/R | White matter lesions >3mm on T2 images |
N/R |
| White matter lesion quantification |
N/R | Volume of white matter intensities measured at baseline. Distribution similar between subject groups |
N/R | N/R | N/R | N/R | N/R | N/R | N/R |
| APOE genotype (% E4) |
N/R | N/R | AD = 54 | AD-C = 54, AD T = 65 | C = 10, slow = 55, fast = 56 |
N/R | C = 22, mild AD =59, moderate=71 |
N/R | N/R |
| Gender (% male) | C = 47, AD = 39 | C = 44, AD = 40 | AD = 40 | AD-C = 19, AD-T = 22 | C = 42, slow = 39, fast = 42 |
AD-C =40b, AD-T = 41b |
C = 48, mild = 52, moderate = 71 |
C = 50, AD = 41 | C = 46, AD = 61 |
| Treatments | N/R | N/R | Donepezil 60%, galantamine 5%, rivastigmine 21%, HRT 21%, Vit E 32% |
Donepezil (AD-T). AD-C, Both groups allowed Vit E, non-steroidal anti-inflammatory, ginko biloba, lecithin |
N/R | N/R | N/R | N/R | N/R |
| Drop-out if RCT | N/A | N/A | 15 due to no post-baseline scan |
N/A | N/A | N/R | N/A | N/A | N/A |
| Assessment of controls |
N/R | MRI excluding neuropathol. |
N/A | N/A | Normal neurological exam, Community dwelling |
N/A | Yearly assessment of neuropsychol., physical and medical exam |
N/R | CDR 0.0 |
| Exclusion criteria for controls |
N/R | Clinical history of alcoholism, psychiatric illness, epilepsy, hypertension, diabetes, major heart disease or head trauma (Du et al., 2003) |
N/A | N/A | No active neurological disorder, psychoactive medication |
N/A | As for AD, plus diabetes mellitus, hypertension, angina, arrhythmia, seizure disorder, cancer, COPD or renal diseasea |
Same as AD | Other neuropsychiatric disorders, genetic family history of AD |
| Magnet strength (Tesla) |
1.5 | 1.5 | 1.5 | 1.5 | 1.5 | 0.5–1.5 | 1.5 | 2 | 1.5 |
| Sites | N/R | 1 | 17 | 1 | N/R | 38 | 1 | 1 | 1 |
| Protocol (thickness (mm)) |
1.5 | 1.4 | 1.5–1.8 | 1.5 | 1.6 | 1.6 | 4 (detailed in Kaye et al., 1997) | N/R | 1 |
| Measurement method Number of raters |
Manual 1 |
Semi-automated Checked by 1 |
Manual 3a |
Manual 1 |
Manual 1 |
Manual 1 |
Manual 3a |
Manual 1 |
Semi-automated Team of experts for template as in Haller et al. (1997), N/R for landmark placement |
| Registered scans or TIV corrected |
Registered | N/R | Registered | N/R | Registered | Registered as referenced in Jack et al. (1998) |
TIV correcteda | Registered | N/R |
| Whole Hc measured? | No (small part of tail excluded) |
No (small part of tail excluded) as in Hsu et al. (2002) |
No (small part of tail excluded) |
No (white matter excluded), Tail excluded |
No (small part of tail excluded) |
No (small part of tail excluded) |
No (body of hippocampus only as in Kaye et al., 1997) |
No (small part of tail excluded) |
No (exclusion of white matter of alveus and fimbria some of tail excluded) |
| Reliability, inter-/intra-rater, same/different acquisitions, repeated × n, number subjects, volume/rate |
N/R | ICC 0.94, intra-rater different acquisitions. Repeated × 2 (assumed), N = 9, volume |
N/R | ICC 0.98, CV 2.8%c, intra-rater, same acquisition. Repeated × 3, N = 10, volume |
ICC 0.91, IQR of difference in rate: 2.3 × 10−3, intra-rater, same acquisition. Repeated × 2, N = 18, rate |
CV 0.28%c, intra-rater, different acquisitions. Repeated × 2, N= 10, volume |
ICC 0.90, intra-rater, same acquisition. Repeated × 5, N= 5, volume |
ICC 0.99 ICC 0.98, intra- and inter-rater, respectively. Same acquisition. Repeated × 2, N= 10, volume |
ICC 0.93, intra-rater (assumed), different acquisitions. Repeated × 2, N= 8, volume |
| Blinding of raters to diagnosis |
Yes | Visually checked blinded to diagnosis |
Blinded to treatment group |
Blinded to AD-C/AC-T | Yes | Blinded to treatment group |
Yesa | Yes | N/R |
| Blinding of raters to order of scans |
Yes | N/R | Yes | Yes | Yes | Yes | Yesa | Yes | N/R |
Key: AD, Alzheimer’s disease; C, control; DAT, dementia of Alzheimer type, ICC, intra-class correlation coefficient, CV, coefficient of variation; AD-C, control group of AD patients; AD-T, treated AD patients; APC, annualised percentage change; N/A, not applicable; N/R, not reported; CDR, clinical dementia rating; MMSE, mini mental state examination; neuropathol., neuropathologies; neuropsychol., neuropsychologies, COPD, chronic obstructive pulmonary disease; slow, slowly progressing AD; fast, quickly progressing AD; mild, mild AD; moderate, moderate AD.
a
Conveyed by personal communication.
b
Based on a wider group.
c
CV quoted by Jack et al. (2003) is calculated as a median (range 0.02–0.70%). CV quoted by Hashimoto 2005 was quoted for left and right hippocampus separately (right = 2.84% and left = 2.78%). This was calculated as standard deviation/mean, where standard deviation indicated square root value of the arithmetic mean of 10 variance estimates.