Association of Low Protein Z with Ischemic Stroke in Young Women.

Protein Z enhances inhibition of factor Xa by plasma protein Z-dependent protease inhibitor (ZPI). Several groups reported low protein Z plasma levels associated with ischemic stroke, but others found high protein Z in association with stroke. We sought to explain why our finding of low plasma protein Z in association with stroke in males (Heeb et al 2002) contrasted with some reports. We found association of low protein Z with stroke in all subgroups (including non-Hispanics) except in smokers, diabetic individuals, and women. The findings were similar in samples taken from stroke patients within 4 days of the stroke, as well as in samples taken 2 months later. Our study differed from others in that our controls were subjects with similar proportions as the stroke patients of known or possible risk factors: hypertension, hyperlipidemia, diabetes, smoking, age and gender, and 61% of our subjects were Hispanic as compared to largely European populations that were otherwise studied. Our proportion of smokers (23%) was smaller than that of European studies and could explain some of the differences in reports. High triglycerides in our controls were significantly associated with high protein Z (p=0.015). Thus, healthy controls used in other studies would be expected to include fewer subjects with high triglycerides and higher protein Z than in our studies. Most other studies did not break patients into subgroups by sex and age. Our study found a strong association of low protein Z with stroke in males (p=0.0004), but only a trend in females (p=0.13). Some studies may have missed these associations because men, nonsmokers, and nondiabetic subjects were not evaluated separately. The lack of significant association of low protein Z with stroke in women could be due to hormonal influences that vary with age. Therefore, we divided our female subjects into ages 24 to 56 and ages 57 to 90. Remarkably, the younger female stroke patients had a significant association of low protein Z levels with stroke compared to controls (mean plasma levels 2.08 versus 2.60 μg/ml, p=0.044), while the older female stroke patients did not (2.22 versus 2.18 μg/ml, p=0.823). Protein Z levels for the younger female cases and controls were nearly identical to those for men in the same age range (younger men: 2.07 versus 2.58 μg/ml, p= 0.037). Older female controls, as compared to the younger female and male controls, had significantly lower mean protein Z (p=0.023), while the older female stroke patients had significantly higher protein Z values than the older male stroke patients (2.22 versus 1.75 μg/ml, p=0.010). Thus, it is difficult to further explain why only the older female stroke patients failed to have a significantly lower mean protein Z level than their control group, even when diabetic individuals were excluded. There was not a lower proportion of older female controls with high triglycerides. In summary, these data show that subgroup analysis by age, gender, and diabetic status is important for plasma protein Z studies and that low plasma protein Z levels are found in association with stroke in males of all ages and in young women, though the association remains strongest in men over age 56 (p=0.002).