Abstract
Purpose: :
To report the relationship between macular pigment optical density (MPOD), C-reactive protein (CRP), and serum levels of lutein and zeaxanthin (L&Z) in elderly subjects.
Methods: :
We studied 249 subjects participating in ARMA, an ancillary study to Health ABC [mean age +/- SD: 78.8 +/- 3.3 yo, 84% White, 55% females, 37% lutein supplement users (LSUs)], inclusive of both subjects with and without age-related maculopathy (ARM), since we have shown that having ARM does not per se affect MPOD [Iannaccone 2007; ARVO E-Abstract 2147]. MPOD estimates were obtained with a previously reported instrument and protocol [Iannaccone et al. IOVS 2007; 8: 1458-65] at 0.5 degrees of eccentricity from the fovea. Fasting serum samples were collected at time of visit and analyzed according to standardized protocols. L&Z and CRP were not normally distributed and were, therefore, ln-transformed.
Results: :
Serum ln(L&Z) levels were strongly correlated to MPOD. However, the correlation was mainly driven by the lowest and highest MPOD quartiles. Serum L&Z displayed no relationship with MPOD values within the inter-quartile range. Likewise, serum ln(CRP) levels displayed a strong inverse correlation with MPOD quartiles, but this relationship was mainly driven by the ln(CRP) levels in the first MPOD quartile. No correlation was present across the subsequent three MPOD quartiles. Lastly, consistent with data previously reported by our group [Kritchevsky et al. AJE 2000; 152: 1065-71; Gruber et al. J Nutr. 2004; 134: 2387-94], serum ln(L&Z) displayed a strong inverse correlation with ln(CRP) quartiles, showing a progressive but curvilinear decline of serum ln(L&Z) levels by ln(CRP) quartile. The relationship with serum ln(L&Z) across the top three ln(CRP) quartiles was much weaker.
Conclusions: :
In elderly individuals, MPOD, ln(L&Z) and ln(CRP) are strongly correlated to one another. However, the relationship is complex and not linear across quartiles. These findings provide an explanation for the variable findings from previous reports, indicate that studies on these variables are likely to show different degrees of correlation between them depending on the range of MPOD, ln(L&Z), and ln(CRP) of the population sample examined, and help us gain insights on the effect of serological variables on the biology of retinal carotenoids.
Keywords: clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • carotenoids/carotenoid binding proteins • inflammation