Analysis of GPx concentration in AMI patients vs. controls

Erythrocite lysate

Results from Tho (1987), Dubois-Rande (1994), and Buko (2014) have been excluded since GPx activity was measured on a “per protein” and not “per Hb” scale.

Values in Díaz-Araya et al. (2002) were multiplied by 100 because the published result was evidently mis-scaled (figure shows ~40 U/gHb the text reports 0.4 nmol).

For Hassanzadeh (2006) SD has been estimated as 8 for CVD patients and 7.5 for Controls.

Meta-regression (GPx activity in Erythrocyte SMD vs year of pubblication)

Characteristic Beta 95% CI p-value
intercept 329 0.26, 657 0.050
year -0.16 -0.33, 0.00 0.050
timeAfter admission -2.7 -5.2, -0.12 0.042
Abbreviation: CI = Confidence Interval
ML REML
ll -25.43621 -21.07784
dev 60.20634 42.15567
AIC 58.87241 50.15567
BIC 61.13221 51.36601
AICc 63.87241 58.15567

Plasma

Results from Dubois-Rande (1994), and Mehri (2023) have been excluded since GPx activity was measured on a “per protein” and not “per Volume” scale.

Results from Wang (1981), and Garelnabi (2010) have been excluded since GPx activity was measured in platelets.

Results from Hazini (2015) have been excluded since GPx activity was measured in erythrocyte and results not expressed “per Hb” (so was excluded from the erythrocyte studies as well).

For Mihailović (2003) results were multiplied by a factor of 60 since expressed in µkat/L. SD has been estimated as 1 for CVD patients and 1.15 for Controls looking at paper figure (https://plotdigitizer.com).

For Bor (1999), and Lafont (1996) results have been multiplied by a factor of 1000 and 10, since expressed as U/ml and U/dl.

Serum

Serum analysis is not included since the results of the three studies selected cannot be safely pooled:

** Akyol (1993) has no health control group (AMI vs Angina): measure unit U/ml; ** Domínguez-Rodríguez (2002) measures the GPx concentration, not the activity; ** Siddiqui (2014) activity expressed “per protein”.

Whole Blood

Results from Mihailović (2003), and Madole (2015) have been excluded since GPx activity was measured on a “per Volume” and not “per gr of Hb” scale (it is eventually possible to pool results from these two studies).

Results from Freitas (2014) has been excluded since GPx activity was measured on a “per protein” scale.

Analysis of GPx concentration in CHF patients vs. controls

Erythrocite lysate

Results from Klimiuk (2020) have been excluded since GPx activity was measured on a “per protein” and not “per gr of Hb” scale.
For Pérez (2002) All results were multiplied by 1000 since reported as U/mgHb. CHF patients data were derived using paper figure (https://plotdigitizer.com).

Plasma

For Mihailović (1998) results were multiplied by a factor of 60 since expressed in µkat/L.

For Keith (1998), and Polidori (2004) results have been multiplied by a factor of 1000, since expressed as U/ml.

Serum

Results from Ping (2023) have been excluded since GPx activity was measured on a “per protein” and not “per Volume” scale.
Liu (2012) results have been multiplied by a factor of 1000, since expressed as U/ml.

Funnel Plot