INTRODUCTION
Aging is associated to a number of changes in the structure and function of various organs.1 In the kidney, among otherchanges, the glomerular filtration rate (GFR) decreases at an average rate of 0.8 mL/min/year.2 It is known that a marked decrease in GFR leads to the occurrence of various manifestations associated to chronic renal failure (CRF), including anaemia3, osteodistrophy4, metabolic acidosis5 and so on.
On the other hand, a great interest is currently being paid to early detection of «occult kidney disease» in the general population from GFR estimation. For this purpose, a number of mathematical equations based on serum creatinine (SCr) have been devised. Thus, depending on GFR, chronic kidney disease (CKD) is divided into several stages. AGFR ranging from 60 and 30 mL/min is considered indicative of stage 3 CKD.6 This classification has been considered applicable to the whole population.
In a prior study,7 we reported that all elderly patients enrolled, including those with SCr values within the normal range, already had decreased GFR values, and approximately one third were in stage 3. Today, a laboratory finding of a decreased GFR in asymptomatic elderly patients with no other associated signs of CKD (anaemia, etc.) is a reason for referral to the nephrology outpatient clinic. This study was therefore intended to establish whether such classification has a practical value in this specific group of patients or is only imposing an unnecessary overload on nephrology clinics. For this purpose, a clinical a laboratory monitoring was performed in a group of patients over 65 years of age, analysing the changes in their kidney function and their morbidity and mortality for one year.
PATIENTS AND METHODS
Patients
Eighty patients > 65 years of age (mean, 82.4 ± 6.5 years; range, 69-97 years) seen during the January-April 2006 period were enrolled. Of these, 38 patients were recruited at the geriatrics outpatient clinic (Group 1, SCr £ 1.1 mg/dL (range 0.7-1.1) and 42 patients at the general nephrology outpatient clinic (group 2, SCr > 1.1 mg/dL (range 1.2-3). Overall, 68.8% of patients were females (84.2% in group 1 and 54.8% in group 2, p < 0.005). A history of diabetes mellitus (DM) and high blood pressure (HBP) was found in 37.3% and 81.3% of patients respectively. Comparison of both groups showed no significant differences in DM or HBP distribution. Study patients were clinically stable and underwent a clinical and laboratory re-evaluation one year later.
Patient distribution by CKD stage at baseline using the Cockcroft-Gault formula was as follows: stage 1, 0%; stage 2, 7.9%; stage 3, 66.6%; stage 4, 20.6%; stage 5, 4.7%. The distribution obtained with the abbreviated MDRD formula was: stage 1, 0%; stage 2, 30%; stage 3, 60%; stage 4, 10%; stage 5, 0%.
Established treatments with statins, antihypertensives, calcium salts, iron, and erythropoietin (EPO).
Methods
A prospective, observational study. The baseline assessment was performed at a scheduled patient visit to the clinic during January-April 2006. All patients underwent clinical and laboratory monitoring for one year, and a re-evaluation was done in the January-April period of 2007. Hospital admissions and their reasons were recorded, as well as the occurrence of cardiovascular events and mortality. Laboratory monitoring was based on GFR estimation using the Cockcroft-Gault8 and abbreviated MDRD formulas.9
Laboratory tests were performed one week before patients attended the scheduled visits at the geriatrics and nephrology clinics both at baseline and at one year. The following parameters were measured in venous blood using standard hospital methods: creatinine, urea, uric acid, lipid profile, albumin, calcium, phosphorus, alkaline phosphatase, electrolytes, haemoglobin, and haematocrit. In group 2, 24-hour urine protein was measured.
Statistical analysis
SPSS 11.0 software was used for statistical analysis. Data are given as proportions, means, and standard deviations. A lineal model for repeated measures was used to assess change in kidney function over time. The significance level was 95%.
RESULTS
Ten percent of patients died before the end of the study year (four due to impairment in overall status, two from fracture complications, one from a tumour, and one from a stroke). Hospital admission of 21.8% of patients was required during the year; 27.8% of these admissions were for cardiac disease, and 22.2% for infections. No significant differences in mortality, admission, and cardiovascular events were found between the groups.
As regards use of drugs, 27.7% were receiving statins, 82.8% antihypertensive drugs, 10.3% calcium salts, and 14.5% iron. No significant differences were found between the groups in use of these drugs. Only some group 2 patients (23%) were receiving EPO (p < 0.003). These were older patients (89 ± 5 vs 81 ± 6 years, p < 0.001) with higher SCr levels (2.0 ± 0.5 vs 1.2 ± 0.2 mg/dL; p 0.000) and lower eGFRs according to the MDRD formula (32 ± 10 vs 54 ± 15 mL/min; p 0.000).
Figure 1 shows the change in kidney function (KF) at one year in the 72 patients who completed the study. While significant differences exist in SCr levels and GFR between both groups (p = 0.000), it may be seen that both SCr levels and GFR (estimated by MDRD and Cockcroft-Gault) do not significantly differ from baseline values.
In group 2, proteinuria showed no significant changes at one year: baseline, 0.14 ± 0.23 g/24 hours (range 0-3 g/24 h), vs 1 year, 0.23 ± 0.46 g/24 h (range 0-1.96 g/24 h).
Figure 2 shows the change in anaemia over time by group. All patients had a haematocrit value higher than 35% both at baseline and at one year of follow-up. No significant differences were found between the groups in all other laboratory parameters tested at baseline and one year.
Similarly, no significant changes over time were found when GFR values were compared between diabetic and nondiabetic patients (table I).
DISCUSSION
There are few studies reporting the progression of CKD in the elderly population. One such report is the Hemmelgarm study,10 describing CKD progression during two years of followup in a population > 66 years. Progression was slow, except in diabetic patients and in those with a GFR < 30 mL/min, in agreement with other studies11,12 where association of estimated GFR to mortality, as well as progression to endstage CRF, was seen to be lower in the elderly population as compared to young patients.
With the current CKD classification, most patients diagnosed the condition are elderly subjects.13 However, only a small proportion of them will require renal replacement therapy, often dying before from cardiovascular disease.14 This study assessed the clinical and laboratory changes over one year in patients over 65 years of age with different grades of baseline GFR. A significant proportion of patients had cardiovascular risk factors (HBP and/or DM), and cardiac disease was the main morbidity cause among the studied groups. One may therefore wonder what benefits may these patients derive from referral to nephrology clinics, whether emphasis should rather be placed on cardiovascular prevention, and where and how would their monitoring be most adequate.
Aging is associated to a decrease in GFR.2 In a previous study,7 our group found that despite the fact that all elderly patients studied had a decreased GFR (even with SCr within the normal laboratory ranges), no characteristic signs of CRF were seen in most of them. Moreover, SCr showed higher statistical correlation levels with the laboratory changes associated to CRF than the GFRs estimated from mathematical formulas. That study concluded that GFR estimations provided no advantages over a simple SCr measurement for assessing KF in the elderly. Moreover, since part of the GFR decrease in these patients is associated to age and the main cause of mortality is cardiovascular disease, the current classification of CKD may induce unnecessary confusion and alarm and does not appear to have practical advantages for management of this type of patients.
In this study, we performed a clinical and laboratory followup to ascertain what happened with GFR. It was seen that, though significant variations existed in GFR levels in both groups, when compared after one year of follow-up, no differences were found, irrespective of such baseline levels.
With GFR decrease, various laboratory changes associated to CRF occur (anaemia, changes in bone and mineral metabolism, metabolic acidosis, etc.). When such changes were looked for, the group of patients with SCr within the normal laboratory range (group 1, SCr < = 1.1 mg/dL), despite having a decreased GFR, showed a baseline haematocrit higher than 35%, that remained one year later at similar levels without the need for using agents to stimulate erythropoiesis. Group 2 patients (SCr > 1.1 mg/dL) were also seen to have a baseline haematocrit higher than 35% that was maintained one year later. The proportion of patients receiving erythropoietin therapy was low. No differences were found between the groups in all other laboratory parameters studied. When the parameters associated to use of erythropoietin in these elderly patients were analysed, SCr levels > 2 mg/dL and advanced age were found to be the two factors that appear to contribute to the need for EPO therapy to prevent the occurrence of anaemia associated to CRF.
It therefore appears that, in standard clinical practice, CKD detection and management in elderly patients could continue to be performed as done to date, in a simpler form, with a simple measurement of SCr and a systematic urine analysis (Multistix) to rule out the presence of kidney disease requiring specialised care. This approach would allow for dispelling the confusion felt by many physicians who are not used to CKD management when they see this type of patients.15 It would also allow for avoiding or reducing unnecessary referrals and overload for the nephrology teams.16 We therefore postulate that elderly patients with SCr higher than 1.1 mg/dL but no changes secondary to CRF (anaemia, etc.) or proteinuria need not be routinely referred to specialised care. In fact, the recommendations recently published in the Consensus Document on CKD of the SEN and the Society of Family and Community Medicine state that patients over 70 years of age with a stable GFR > 30 mL/min and albuminuria < 0.5 g/24 h may be monitored at primary care without the need for referral, provided adequate control is maintained of blood pressure and all other vascular risk factors (recommendation 11).17
In conclusion, in elderly patients showing no significant proteinuria or laboratory changes associated to CKD (anaemia), the similar clinical course (irrespective of baseline GFR) and the stability of the various laboratory parameters associated to the decrease in kidney function over time allow for simplification of follow-up, with no need to resort to GFR estimation using mathematical formulas.
