Demographic and clinical findings
The search process enabled the inclusion of 194 patients into the study, and 97 patients were in each group. Patients in the allopurinol group were matched to control group for age and sex, and for mean achieved blood pressure and heart rate: respectively, 133/79, 80 vs 132/79mmHg, 80 beats/min). The allopurinol group had significantly higher body weight and body mass index (BMI) (estimated differences of, respectively, 7 Kg (CI; 1, 13, P= 0.020) and 2.1 Kg/m2 (CI; 0.2, 4, P= 0.029). In terms of the co-morbid conditions, the two groups were well matched (see Table 1) apart from the prevalence of a history of systemic hypertension and chronic kidney disease which were significantly higher in allopurinol group; respectively, 85.6% vs 67% (P=0.002) and 36.1% vs 7.2% (P<0.001) (Table 1). Further classification of the patients based on estimated GFR (calculated by the abbreviated MDRD equation:
186 x (Creatinine/88.4)-1.154 x (Age)-0.203 x (0.742 if female) x (1.210 if black).
In general, across the stages, renal impairment was more prevalent in the allopurinol group. This included 12 patients in the allopurinol group (vs none in the control group) who were maintained on renal replacement therapy: 2 on hemodialysis and 10 on peritoneal dialysis.
Allopurinol indications and dose
Allopurinol therapy with indication, dose, and duration (in months before HbA1c readings) are listed in table 1.The most frequent dose was 100 mg/day in 77.3% of the whole group, while the median duration for allopurinol therapy was 39.5 months (10.95, 82.78)
Comparisons of parameters of glycemic control (GC)
HbA1c readings were compared between the two groups and showed a small yet significant difference in favor of allopurinol therapy. The estimated difference was -0.5%, (CI: -0.8, -0.1) (P=0.010)(Table 1).
Sensitivity analysis
Advanced chronic kidney disease (CKD) is associated with shortened red blood cells lifespan which might reduce HbA1c readings. Accordingly, the HbA1ccomparison was repeated after removal of the 12 patients on renal replacement therapy (85 allopurinol patients’ vs 97 control patients). The estimated difference obtained for HbA1c was -0.4%, (CI: -0.7, -0.0) (P=0.033).
Comparisons of parameters of glycemic variability (GV)
- Comparing FBS readings between the two groups is demonstrated in (Table 1) as medians with interquartile ranges (Q1, Q3) with a trend towards lower reading in favor of allopurinol users, with an estimated difference of -8 mg/dL (CI: -17.33, 1) (P=0.072).
- Comparing the range of FBS readings with the minimum and maximum values were; 84.33, and 287.00 mg/dL in the allopurinol group, as compared with 57.00, and 372.00 mg/dL in the control group (See Fig. 2).
- Comparisons of standard deviation, and coefficient of variation showed a trend for less dispersion in FBS readings in allopurinol users (Table 1).
- Comparisons of the VIM showed a statistically significant results in favor of allopurinol therapy (P<0.001) (Table 1 and Figure 3).
Correlation between GC and GV indices
By studying the correlation between two parameters of glycemic control; HbA1c, and FBS, with an index of GV; VIM, correlation analysis revealed that FBS was not correlated with VIM in both groups; Spearman correlation coefficients (rho); (r=-0.018, P=0.876) in allopurinol group, and (r=0.027, P=0.816) in control group. While HbA1c had no correlation with VIM in allopurinol group (r=0.044, P=0.701), it had a significant positive correlation in control group (r=0.273, P=0.015).
Comparisons of other biochemical parameters
As expected, the renal biochemical parameters (blood urea nitrogen, serum creatinine and serum uric acid concentrations) were higher in the allopurinol group (Table 1). There were no clinically relevant differences for sodium, potassium or chloride.
Comparisons of pharmacological therapy
The two groups were matched for diabetes mellitus medications but they differed slightly in the use of angiotensin receptor blockers (52.6% in the allopurinol group vs 36% in the control group), beta blockers and calcium channel blockers (49.5% vs 27.8%). There also were small differences for diuretic and aspirin use (Table 2).