To the Editor:
We thank Dr. Agilli and colleagues1 for their interest in our study of 217 patients with rheumatoid arthritis (RA)2. This investigation documented an independent relationship of leptin concentrations with carotid artery plaque in those with but not without a large cardiovascular disease (CVD) risk burden as represented by the presence of traditional risk factors2.
In our data analysis, we took into account the potential confounding or/and mediating effect of patient characteristics including a range of demographic features, lifestyle factors, anthropometric measures, RA characteristics, systemic inflammation, and kidney function. Interestingly, we found that the leptin concentration/atherosclerosis relationship was driven by the glomerular filtration rate and body mass index2. The included potential confounding or/and mediating variables in our analysis had been identified in a previous study that also showed an independent association of leptin concentrations with surrogate markers of early atherogenesis among young patients with RA in the present cohort3.
Dr. Agilli and colleagues point out that, based on recently reported evidence, an additional 18 patient characteristics1 can influence circulating leptin levels, and therefore may have required consideration in our analysis. These include comorbid conditions such as major depression, chronic liver diseases, systemic lupus erythematosus (SLE), psoriasis, multiple sclerosis, and Helicobacter pylori infection, the use of glucocorticoids, and antipsychotic, antihypertensive, lipid-lowering, hormonal, insulin sensitivity–altering, and antidepressant medications, and dietary supplements such as vitamins A, D, and E, linoleic acid, and omega-3 fatty acids1.
Among our participants, none had concurrent chronic liver disease, SLE, psoriasis, or multiple sclerosis. H. pylori status was not assessed. None of the included patients used antipsychotic agents, and vitamins A and E, linoleic acid, and omega-3 fatty acids were not prescribed by treating physicians.
Data on the remaining 8 potential confounding or/and mediating characteristics1 that could have been relevant in our analysis were recorded. These are given in Table 1. In this regard and as part of our routine practice, we systematically recorded the Arthritis Impact Measure Scales (AIMS) depression score4, because this is highly relevant in the present context4,5. Among patients with a large traditional CVD risk burden in the present study, the mean (SD) AIMS depression score was 3.0 (1.8), and 32.1% experienced a score of > 4, which can indicate the presence of clinical depression4. Apart from depression, the other potential confounding or/and mediating characteristics were the use of medications or groups of medications. This included vitamin D supplementation that, at the time of our study, formed part of treatment regimens for osteoporosis, which included a bisphosphonate or strontium ranelate. Notably, vitamin D is implicated in CVD risk in RA6,7.
As also shown in Table 1, among the 8 respective characteristics1, an AIMS depression score of > 4 as well as the use of potential glucose sensitivity–altering medications were in fact associated with leptin concentrations in an analysis adjusted for demographic variables. This does support the issues raised by Dr. Agilli and colleagues1.
In our previously reported study2, we found that a 1-SD increment in leptin concentration increased the OR for plaque 2.75-fold (95% CI 1.19 to 6.37; p = 0.01) after adjustment for potential confounding or/and mediating characteristics in a mixed regression model. Glucocorticoid therapy was used by only 1.9% of the participating patients with a large CVD risk burden, and the effect of glucocorticoid therapy on the leptin concentration/atherosclerosis relationship was not assessed, because none of the patients using the respective intervention had plaque. When we reevaluated the leptin concentration/carotid artery plaque relationship with the additional inclusion of any of the remaining 7 newly identified potential confounders or/and mediators including depression or any of the recorded interventions in the mixed regression model2, the respective association persisted on each occasion (Table 2).
It is possible if not likely that some of our patients were using dietary supplements without our knowledge. Despite this, taken together, the comments by Dr. Agilli, et al1 and the results of the ensuing analysis as presented here strengthen our finding that leptin can contribute to atherogenesis, and consideration of circulating concentrations of this adipokine may be useful in CVD risk stratification among patients with RA2,3.
Footnotes
-
Supported by The South African Medical Research Council (grant MRC2008_DES) and the National Research Foundation.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.