Back from a meeting, I can summarize a few reasons the Sage journal article likely should not have been published as framed.
The foundational problem is that this a post-hoc analysis which, especially with this weak sub-design, is not powered to illuminate the klotho issue. The original crossover trial (Lim et al., 2007) was designed to compare losartan versus quinapril for albuminuria reduction in Asian T2DM. Klotho was measured retrospectively in stored samples, or added as a secondary endpoint without prospective power calculation. This means that no a priori sample size calculation for the Klotho endpoint exists and the risk of Type I error inflation from multiple post-hoc comparisons is uncontrolled. One can stop here. At best, this is an hypothesis-generating study.
Irrespective of design issues, the study is significantly underpowered. Thirty-three subjects in a crossover design means 33 observations per arm. For a biomarker with known high inter-individual and intra-individual variability (as I suspect all in this area are) detecting a meaningful between-treatment difference requires substantially larger samples. (In one search, I estimated that klotho has a coefficient of variation in healthy populations of more than 20%.)
It is obvious but noteworthy that there no placebo arm. The study compares losartan to quinapril absent an untreated control, a weakness in the original study compounded here. This means the study can tell you, at best, that losartan raises Klotho relative to quinapril (or that quinapril suppresses Klotho relative to losartan) or that both raise Klotho from baseline but at different rates. You cannot determine from this design whether losartan raises Klotho above untreated baseline. IMO, this is not a minor issue.
I’m sure everyone reading this paper wondered why both drugs were administered at 50% of their maximal approved doses. This is a non-standard approach. Low dose and short time lines are what I often see in a study that wants to make a preconceived point. Although I don’t think that is the case here, it might be the case that the timeline was “shopped” for significance.
I don’t know much about klotho but four weeks seems short to claim a steady state. I have seen ARB studies spanning 24 weeks. Perhaps an expert can add to that.
Another flaw that would benefit from the opinion of an expert is that T2DM patients with pre-existing albuminuria have structurally compromised kidneys, which is the primary organ of Klotho production. It would seem to me, then, that this population likely behaves differently from those in healthy hypertensive individuals. The generalization back to the population seems like an unsubstantiated leap.
Finally, my initial irritation, the title claim: “associated with reduction of albuminuria.” The title asserts an association between losartan-induced Klotho elevation and albuminuria reduction. This is a correlation between two treatment-responsive variables in a four-week trial. It cannot cannot establish that Klotho elevation mediates the anti-albuminuric effect. This problem could have been remedied with a more thoughtful title.