In order to assess the diagnostic LY2157299 utility of cystatin C, we first compared the mean first-draw cystatin C levels among ALS patients, neurologic disease controls, and LY2835219 CDK inhibitor healthy controls. A generalized linear model was used to estimate the mean total and percent cystatin C for each diagnostic category, with both gender and age included as co-factors in the model. This statistical design controls for between-group differences in each co-factor when generating estimated means. Therefore, the differences in age and gender among our diagnostic groups should not have affected our results, even in the case that cystatin C varies with these factors. We found that the estimated means for both measures of cystatin C were lower in ALS patients than in disease controls and healthy controls, similar to prior studies. However, a test of the model��s main effects revealed that only percent cystatin C differed significantly by disease diagnosis, while total cystatin C levels were not significantly different across diagnostic groups. Neither measure of cystatin C differed significantly by age or gender. In a post-hoc pairwise comparison of diagnostic groups, percent cystatin C was found to be significantly lower in CSF of both ALS patients and disease controls relative to healthy controls, but there was no statistical difference between cystatin C levels in ALS patients and disease controls. Next, we repeated these statistical analyses using data from specific patient subgroups, in order to determine if either measure of CSF cystatin C can be used to differentiate ALS patients from disease controls in specific patient subpopulations. First, we created a subcategory of disease controls comprised of patients with neurologic diseases that more closely resemble ALS at presentation. Using this group of ALS mimics in our statistical model, the patterns of overall and between-group statistical differences remained the same. However, the p-values were reduced for the ALS vs. mimic disease control subgroup comparison relative to the ALS vs. all disease control subgroup comparison suggesting a stronger trend toward statistical significance when cystatin C is used to differentiate ALS patients from this more clinically-relevant control group. We next compared two ALS subgroups to the disease mimic group. Limb-onset ALS and ALS patients greater than one year from symptom onset both exhibited reduced levels of cystatin C in the CSF when compared to disease mimics, with improved p-values when compared to the analysis including all ALS patients. However, the pair-wise comparisons between these ALS subgroups and mimic disease controls still fell short of statistical significance.