Glomerular filtration rate and serum phosphate: an inverse relationship diluted by age

Methods. Data were collected on age, sex, menstrual status, anthropometry, overnight urinary creatinine, dietary protein (overnight urinary urea), reported intake of milk/yogurt, serum creatinine, phosphate, calcium and total protein in 4034 adults (age 18-91 years) with GFR >= 60 mL/min x 1.73 m(2) as assessed by estimated GFR (eGFR, simplified MDRD equation) and creatinine clearance (overnight urinary creatinine/serum creatinine). Results. The relationship of eGFR with serum phosphate was positive in men and null in women in univariate analyses (P = 0.001 and 0.148), negative in both sexes with age adjustment (P < 0.001). Age-adjusted results did not depend on colinearity between age and eGFR because the relationship was inverse also replacing eGFR with creatinine clearance (P < 0.001 in both sexes). In univariate regression analysis done separately by gender and six age-strata (18-24, 25-34, 35-44, 45-54, 55-64 and >= 65), the line of serum phosphate over eGFR was constantly inverse (range of P = 0.010/0.089) with the progressively lower y-axis intercept from young to older ages. The inverse relationship of eGFR or creatinine clearance with serum phosphate was significantly inverse also controlling for other variables (P < 0.01). Conclusions. GFR differences in the range >= 60 mL/min x 1.73 m(2) are inversely and independently related to serum phosphate. The relationship is undetectable without age-controlled procedures because, for serum phosphate, the effect of GFR differences above >= 60 mL/min x 1.73 m(2) is much smaller than the effect of age.