We read with great interest the article by Duggan and colleagues who found in a randomized controlled trial that dietary weight loss, with or without exercise, resulted in significantly reduced plasma concentrations of some circulating oxidative stress biomarkers in postmenopausal women after a 12-month period (1). The greatest reduction in the diet (−23%), exercise (−15%), and diet + exercise (−24%) groups was observed for F2-isoprostanes, which were measured by a gas chromatography–mass spectrometry method and include 8-iso-prostaglandin F2α (2). Yet, the utility of 8-iso-prostaglandin F2α, malondialdehyde (MDA) and 4-hydroxy-2-nonenal (HNE) as markers of free radical–induced oxidative stress is limited because they are also produced enzymatically from arachidonic acid by COX (3–5).
A major concern of measuring circulating 8-iso-prostaglandin F2α, MDA, and HNE is their easy and abundant artifactual formation, even if plasma/serum samples have been immediately generated after blood sampling and stored at −80°C. Artifactual formation of circulating 8-iso-prostaglandin F2α, MDA, and HNE is not fully understood. Its prevention is incomplete and has become a serious problem in long-term clinical studies. We performed two placebo-controlled clinical studies on elderly patients with peripheral artery occlusive disease (PAOD) or coronary artery disease (CAD), which ran for 3 (PAOD) or 6 (CAD) months. We observed much lower 8-iso-prostaglandin F2α, MDA, and HNE plasma levels in the samples collected at the end of the studies than in those collected at the beginning of the study, both in the placebo and the verum groups (Fig. 1; refs. 4, 5). Yet, this was not the case when these biomarkers were measured in urine samples collected concomitantly with the blood samples at baseline and after treatment (4, 5). Most likely, this is because urine is regularly poor in lipids. Thus, inclusion of placebo/control groups is important but does not per se correct for artifactual formation of lipid peroxidation biomarkers.
Plasma concentrations of 15(S)-8-iso-prostaglandin F2α (A; 15(S)-8-iso-PGF2a), MDA (B), and HNE (C) in plasma of 40 patients (age, 68 years; 31 males, 9 females) with PAOD at baseline (0 day) and after 90 days (90d) of oral administration of l-arginine (Arg, n = 20) or placebo (Plac, n = 20). The figure was constructed using previously reported data (4, 5).
Plasma concentrations of 15(S)-8-iso-prostaglandin F2α (A; 15(S)-8-iso-PGF2a), MDA (B), and HNE (C) in plasma of 40 patients (age, 68 years; 31 males, 9 females) with PAOD at baseline (0 day) and after 90 days (90d) of oral administration of l-arginine (Arg, n = 20) or placebo (Plac, n = 20). The figure was constructed using previously reported data (4, 5).
In long-term clinical trials on oxidative stress, reporting the age of the study blood/plasma/serum samples at the time point of analysis is essential. Duggan and colleagues reported that paired samples were randomly placed across batches (1). The decrease in plasma F2-isoprostanes observed in that study could be therefore due to the great difference of the samples' age at the time of analysis. The comparably lower extent of decrease seen in the control group (1), which is, however, not useful as a placebo group, could be due to differences in nutrition, for example, ingestion of higher amounts of polyunsaturated fatty acids (6).
We appreciate the contribution of Duggan and colleagues to the potential importance of dietary weight loss, physical exercise, and oxidative stress on postmenopausal women (1). Reporting of additional information from long-term clinical trials as discussed above will help improve our understanding of the nature and significance of oxidative stress in health and disease. It seems that 8-iso-prostaglandin F2α, MDA, and HNE are not useful as lipid peroxidation biomarkers when measured in blood/plasma/serum samples.
See the Response,
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
RSS Feeds