We appreciate the interest and comments about our article. Drs Tariq and Mitchell state that some literature on Stratus OCT report retinal thinning in diabetics with minimal or no diabetic retinopathy (DR) —in contrast to our article, which reported no thinning compared with controls. Drs Tariq and Mitchell also ask whether such discrepancy between studies may be accounted for by correcting for the effects of race, age, and sex on retinal thickness (as we did in our study). This is an excellent point and, in our opinion, a very important consideration when designing or interpreting studies with OCT-based outcomes.
Our study as well as others have shown that retinal thickness values vary based on age, race, and sex. Many studies do not report detailed demographics of their study groups. In most studies, the distributions of race and sex are skewed because these variables are not commonly thought to effect the outcomes. For example, Biallosterski and associates report a decreasing trend in macular thickness parameters in diabetics versus controls; however, these results may be confounded by the increasing ratio of females in the diabetic versus control group. Similarly, Oshitari and associates found that diabetics without DR had thinner macular measurements and that diabetics with nonproliferative DR had thicker macular measurements than controls; these results also can be confounded by the increasing ratio of males in their diabetic groups. In contrast, Bressler and associates report no difference between diabetics without DR and nondiabetic controls (similar to our study). This finding may be the result of the well-balanced sex ratio as well as the small percentage of nonwhites (19%) in their study. In all cases, it is important to recognize that the differences are small in magnitude (10 to 20 μm) and may need to be confirmed by higher-resolution spectral-domain technology.
Drs Tariq and Mitchell ask whether our study would have shown a difference in macular thickness before we adjusted for age, race, and sex. In response, we have repeated the calculations in Table 1 without adjusting for any demographic variables and presented the results in the appended Table 1A above. Adjustment for age, race, and sex decreased the significance of the difference in several retinal thickness parameters between diabetics and nondiabetics ( Table 1A vs original Table 1 ). Specifically, temporal inner macula, superior inner macula, and superior outer macula are thinner (statistically significant) in diabetics than controls when we do not adjust for age, race, or sex. Without the adjustment, other parameters trend toward significance as well. We could not extend this subanalysis to the ethnic and gender subgroups because the number of subjects in some groups is insufficient for such an analysis. However, the findings of Table 1A confirm the importance of adjusting for demographic variables in future studies.
