meta <- metacont(N.HC, Mean.HC, SD.HC, N.MCI_AD, Mean.MCI_AD, SD.MCI_AD, sm=“SMD”,

data=data, studlab=paste(Author, Year))

## Warning in metacont(N.HC, Mean.HC, SD.HC, N.MCI_AD, Mean.MCI_AD, SD.MCI_AD, :

## Note, studies with non-positive values for n.e and / or n.c get no weight in

## meta-analysis.

meta$label.e <- “HC”

meta$label.c <- “MCI_AD”

print(meta, digits=2)

## SMD 95%-CI %W(fixed) %W(random)

## Bjerke 2009 3.40 [2.63; 4.16] 2.5 6.9

## Mattsson 2009 1.31 [1.13; 1.49] 44.8 10.4

## Hertze 2010 1.54 [1.06; 2.01] 6.4 8.9

## Palmqvist 2012 NA 0.0 0.0

## Spencer 2019 NA 0.0 0.0

## Hansson 2006 2.68 [2.12; 3.25] 4.6 8.3

## Hansson 2007 1.66 [1.18; 2.15] 6.1 8.8

## Hampel 2004 1.95 [1.10; 2.80] 2.0 6.4

## Herukka 2007 1.43 [0.96; 1.90] 6.5 8.9

## Lanari 2009 NA 0.0 0.0

## Papaliagkas 2009 NA 0.0 0.0

## Eckerstrom 2010 0.77 [0.16; 1.38] 3.9 7.9

## Kester 2011 NA 0.0 0.0

## Seppala 2011 1.03 [0.31; 1.75] 2.8 7.2

## Buchhave 2012 2.47 [1.96; 2.98] 5.6 8.6

## Parnetti 2012 NA 0.0 0.0

## Prestia 2013 NA 0.0 0.0

## Leuzy 2015 NA 0.0 0.0

## Baldeiras 2018 1.36 [0.99; 1.74] 10.4 9.5

## Khoonsari 2019 1.51 [0.93; 2.09] 4.3 8.2

## Santangelo 2020 NA 0.0 0.0

## Eckerstrom 2020 NA 0.0 0.0

##

## Number of studies combined: k = 12

##

Number of observations: o = 1855

##

## SMD 95%-CI z p-value

## Fixed effect model 1.53 [1.41; 1.65] 24.86 < 0.0001

## Random effects model 1.73 [1.39; 2.07] 9.98 < 0.0001

##

## Quantifying heterogeneity:

## tau^2 = 0.2799; tau = 0.5291; I^2 = 83.7% [72.9%; 90.1%]; H = 2.47 [1.92; 3.19]

##

## Test of heterogeneity:

## Q d.f. p-value

## 67.35 11 < 0.0001

##

## Details on meta-analytical method:

## - Inverse variance method

## - DerSimonian-Laird estimator for tau^2

## - Hedges’ g (bias corrected standardised mean difference)