Kidney Function in An Unselected Lithium Population

Contact: joanna.jia@wecistanche.com / WhatsApp: 008618081934791

H. Bendz, et al

ABSTRACT 一 We studied kidney function in 124 short-term and long­term lithium outpatients from a population of 127 patients. The glomerular and distal tubular functions were measured and correlated with a num­ber of demographic and treatment variables. There was a significant negative correlation between age and glomerular filtration rate. There were no other significant correlations. The tubular function was below normal in 51 % of the patients. The glomerular function was below nor­mal in 3 % of the patients. We conclude that lithium treatment in a non-toxic dose affects kidney function and that tubular function is more affected than glomerular function. Tubular function probably is better than our figures indicate, glomerular function is not as good. Types of lithium preparation do not affect kidney function differently nor does combined treatment with neuroleptics.

Cistanche for improving kidney function

Since lithium was introduced as a treat­ment for psychiatric disorders, its effect on kidney function has been a matter of concern. Polyuria, polydipsia, and reduced urinary concentrating ability are among the effects that have been noted (1). These effects have been considered reversible (2-5). However, in 1975, Lindop & Pad­field (6) published a case of permanent nephrogenic diabetes insipidus. Hest:bech et al. (7) and Hansen et al. (8) studied a group of lithium patients, admitted for either lithium intoxication or polyuria. The authors expected to find morphological changes of an acute character but found chronic changes instead. This find­ing, suggesting a possibly serious nephro­toxic effect of lithium treatment, attracted worldwide attention and led to several investigations of kidney function.

The present investigation was conducted at the psychiatric clinic, Sahlgrenska Hos­pital, Gothenburg, Sweden, and was aimed at answering the following questions:

1) Does lithium cause a disturbance of kidney function?

2) Is tubular or glomerular function dis­turbed?

3) Is there a connection between such a disturbance and psychiatric diagnosis?

4) Is there a connection between such a disturbance and treatment variables?

5) Does joint neuroleptic treatment affect kidney function?

Material

Patients. The investigation started in March 1978. All 127 patients who were then on lithium were included. One-hun­dred and twenty-four had gone through the program by the end of 1979. Those who didn't were three women: one was manic and left the area. Two refused to participate after the determination of their endogenous creatinine clearance.

Sex and ages. See Table 1.

Psychiatric diagnoses. See Table 2.

Diagnostic criteria. Unipolar affective dis­order (UP) 一 at least three distinct epi­sodes of melancholic or non-melancholic depression. Bipolar affective disorder (BP) 一 at least one complete cycle of depres­sion and mania (hypomania) or at least one episode of mania (9, 10). Cycloid psychosis (CP) 一 criteria according to Per­ris (11). Unspecified affective disorder (US) 一 does not meet the criteria for either UP, BP, or CP. Schizophrenia (SP) -criteria according to Bleuler (12).

The diagnosis refers to the beginning of the lithium treatment. When the course of the illness clearly showed that the origi­nal diagnosis was incorrect it was revised by us.

Somatic disease.

None of the patients ad­mitted to phenacetin consumption. Forty patients reported some kind of urinary or cardiovascular disease. Of these, 15 patients had a history of probably clini­cally significant kidney, cardiac, or vascu­lar disorder. At the time of this investi­gation, six of the patients had a signifi­cant bacteriuria without symptoms of renal involvement. One patient had dia­betes treated with oral hypoglycemic agents, three patients had a diastolic blood pressure of 105 mmHg. One of them was on medication for hypertension before the investigation began. Another four patients were on antihypertensive medication. None had a disturbance of electrolyte balance.

Thyroid function.

Ten patients were tak­ing thyroid substitution medication (levo­thyroxine). Two of these and another five had an increased level of thyroid stimu­lating hormone (TSH).

Maximal 12 h serum lithium.

Five pa­tients had had at least one episode of se­rum lithium > 2.0 meq/L

Cistanche can improve kidney function

Methods

Design

All but one of the 124 patients were ex­amined as outpatients by H.B. or S.A. We interviewed the patients and reviewed the charts with regard to diagnosis, lithi­um treatment, neuroleptics, serum lithi­um, side effects of lithium treatment, cardiovascular and urinary tract disorders in the patient and his relatives, and phe­nacetin consumption. Physical examina­tion included: palpation of the thyroid gland, blood pressure (patient lying), height, and weight. Analyses of blood and urine were done on three separate oc­casions at an interval of 1 week or more. The following analyses were carried out:

1) in blood: serum Hb, electrolytes (in­cluding Ca), lithium, TSH, and creatinine; in urine: protein, glucose, lithium, creati­nine, microscopy, culture, and 24-h vol­ume. 24 h endogenous creatinine clearance was calculated in an ordinary way.

2) DDAVP-test (13). 3) 5lCr-EDTA-clear- ance (14).

Serum creatinine was estimated according to routine laboratory methods.

Creatinine clearance (Ccr) was only per­formed once. Patients collected 24-h urine at home. Satisfactory compliance was pre­sumed in patients who excreted between 70 and 110 % of nominal daily lithium dose in their 24-h urine. Values below 90 ml/min/1.73 m2 were checked by the 51Cr-EDTA-method, when possible. Val­ues above 90 ml/min were checked by the 51Cr-EDTA-method if we suspected labo­ratory error and on patients who were to participate in a lithium withdrawal study.

51Cr-EDTA-clearance was performed with the patient lying. Fifty-seven patients were examined. They did not differ from the whole population regarding sex or age.

Maximal U-osmolality DDAVP-test was performed as an intranasal DDAVP ad­ministration after 12 h (overnight) thirst. Patients took their last lithium tablet be­fore the thirst period. Compliance with the thirsting rule could not be controlled. Osmolality was measured by the freezing point technique in a Roebling machine.

Reduced kidney function. U-osmolality below 800 mmol (DDAVP-test) or glomerular filtration (GFR) below -2 SD for age measured by the slCr-EDTA- method (15).

Statistics

Kidney function values (serum creatinine, endogenous creatinine clearance, 51Cr- EDTA-clearance, urine osmolality) were correlated with sex, age, diagnosis, time on lithium, average and maximal serum lithium during time-on-lithium, the total amount of ingested lithium expressed as the product of time-on-lithium and aver­age serum lithium, type of lithium prepa­ration, neuroleptic treatment and side ef­fects of lithium treatment. Twenty-four-hour urine volume was correlated with urine osmolality. Statistical significance was tested by the Student's t-test and re­gression analysis, covariance analysis, and multiple, stepwise regression analyses. Sig­nificance at the 5%, 1%, and 0.1% levels was designated with one, two, and three asterisks respectively.

Multiple, stepwise regression analysis included as dependent variable the kidney function parameters mentioned above one by one and as independent variable age, average serum lithium, time-on-lithium and the total amount of lithium ingested ex­pressed as the product of time-on-lithium and average serum lithium. The analysis was performed on the whole population (n = 124) and on the following subgroups: all men (n = 46), all women (n = 78), all patients who underwent the 51Cr- EDTA examination (n = 57), and all pa­tients who excreted in their urine between 70 and 110 % of nominal daily lithium dose (n = 69). The subgroups did not dif­fer in results from the whole population, which is, therefore, the only one presented. We performed covariance analysis to examine the effects of treatment with neu­roleptics and type of lithium preparation on renal function; the covariate was age.

This investigation was approved by the ethical committee at the Medical Faculty in Gothenburg.

cistanche male benefits

Results

Renal function

S-creatinine was slightly above normal in five of the 46 male patients (normal is less than 120 /zmol/1). All of these had normal slCr-EDTA-clearance. The aver­age for males was higher than for fe­males (91 and 78 //mol/1 respectively; t = 5 37***)There were no differences be­tween age or diagnostic groups.

Twenty-four-hour endogenous creatinine clearance is missing for one of the 124 patients and was below 90 ml/min in 18 males and 30 females. Forty-three of these 48 patients went through the 51Cr- EDTA examination. Values below normal were found in two females. There were no differences between diagnostic groups.

In 55 % of the patients with endogen­ous creatinine clearance < 90 ml/min, a numerically higher value was recorded on 51Cr-EDTA-clearance. In almost all of the patients with endogenous creatinine clear­ance > 90 ml/min, a numerically lower value was recorded on 51Cr-EDTA-clearance (Table 3).

slCr-EDTA~clearance. There was no dif­ference between diagnostic groups. Values were below normal in two male and two female patients. One man had been worked up for decreased renal function before he was seen by us. He was ex­amined as an inpatient at the nephro­logical department, Sahlgrenska Hospital. Endogenous creatinine clearance and TSH were not done. The second man with a decreased 51Cr-EDTA-clearance had nor­mal creatinine clearance (> 90 ml/min). We studied this man's renal function as part of a lithium withdrawal study.

All four patients had only moderate decreases in GFR. Osmolality was consid­erably reduced in two patients, and little or not at all in two. The patients had been taking lithium for 27 to 80 months. None of them had a history of significant renal disease, lithium intoxication, or increased TSH (see Table 4).

Maximal osmolality. Values were below 800 mosm/kg H2O in 20 male and 43 fe­male patients or 51 % of the population. Between diagnostic groups, there were no differences. Six out of seven hypertensive patients had urine osmolality below 800. All of them had normal GFR. Ten out of 15 patients with significant renal or cardiovascular disease unrelated to lithi­um treatment had urine osmolality below 800. All of them had normal GFR.

Twenty-four-hour urine volume. Fifteen patients were polyuric (above 3,000 ml). Twelve of these had a maximal urine osmolality below 800 compared to 47 out of 106 non-polyuric patients.

Thirst was reported by 74 patients ("thirst patients5,). Their 24-h urine volume (mean =2.6 1) was significantly higher than 37 patients who did not report thirst (mean = 1.8 1; t = 3.96***). All polyuric pa­tients complained of thirst as did some non-polyuric patients. The lowest 24-h urine volume associated with thirst was 800 ml. The maximal 24-h U-volume in non­thirst patients was 2,600 ml. Urine osmo­lality in 79 thirst-patients (mean = 721) was significantly lower than in 40 non­thirst patients (mean = 816; t = 3.02**).

The patients were divided into three groups: 1) reported all the following side effects: thirst, polyuria (subjectively), and frequent urination (subjectively); 2) re­ported none of these side effects; 3) others. There was no difference between groups 1 and 2 in urinary concentration.

Relationship between function and therapy variables

Simple correlations. We found the ex­pected significant, negative correlation be­tween glomerular function and age. No other significant correlations were found. See Table 5.

Relationship between measures of kidney function. We found a significant negative correlation between urine osmolality and volume (r = -0.51**), a significant posi­tive correlation between maximal urine osmolality and 51Cr-EDTA-clearance (r = + 0.29*, partial correlation, correcting for age).

Regression analysis

Multiple, stepwise regression analysis with the four measures of renal function as de­pendent variables showed that age alone contributed significantly to the variance. Age was significantly related to measures of glomerular function, but not to mea­sures of tubular function (Table 6). In general, the common variance was low.

Covariance analysis

Type of lithium preparation. The whole population was divided into three groups, defined as follows: 1) treated only with rapidly dissolving (RD) type during at least 90 % of the time; 2) treated only with slow-release (SR) type during at least 90 % of the time; 3) others. Covari­ance analysis of the first two groups with age as a covariate did not show any differ­ences in the four measures of renal func­tion. Thus, types of lithium preparation seem to have no different effect on renal function. The effect of lithium preparation on osmolality is shown in Fig. 1.

Neuroleptic treatment. The whole popula­tion was divided into three groups, defined as follows: 1) treated with neuroleptics of any kind (except for dixyrazim and alimemazin) and any dosage, for altogether at least 2 years during their whole life­time; only treatment periods of at least 1 week continuously were included; 2) not treated with neuroleptics; treatment de­fined as above; 3) others.

Covariance analysis of the first two groups with age as a covariate did not show any difference in the four measures of renal function. Thus, treatment with neuroleptics seems to have no effect on renal function. The effect of neuroleptic treat­ment on osmolality is shown in Fig. 2

Discussion

Sources of error

We screened our patients with a variety of laboratory tests. In addition, we inter­viewed them. We found that a number of them showed signs of significant kidney disease. Besides that, occult kidney dis­ease does not seem likely to have in­fluenced our results to any substantial de­gree.

We drew blood for serum lithium de­termination 12 h after the last dose of lithium during this study. However, it was clear from the charts that many previous serum lithium levels were not 12-h values. We calculated the mean serum lithium level using all values after the first month of treatment, even though the uncertainty about the timing of the blood draws makes the averages uncertain. Serum lithi­um is determined more frequently in inpatients than in outpatients, so inpatient values got a relatively greater weight in the mean. It is reasonable that this should be the case since inpatient values are more likely to follow the 12-h rule. Lithium levels from laboratories other than Sahlgrenska were included.

We discussed the possibility of esti­mating the total amount of lithium in­gested by using chart records, but patient compliance during many years of treat­ment is difficult to assess. For this reason, the total amount of lithium was defined as: mean serum lithium multiplied by time- on lithium.

Treatment with neuroleptics was quanti­fied only in terms of time. Data on neuro­leptic treatment were more difficult to extract and interpret from the charts than lithium treatment. Time-on-neuroleptics is therefore a more uncertain measure than time-on-lithium. In order to diminish the influence of this uncertainty, we divided the population into two extreme subgroups.

Cistanche can improve kidney function

Results

We did not find any relationship between kidney function and the following patient and treatment characteristics: sex, diag­nosis, time-on-lithium, average serum lithium, the product of the latter two, maximum serum lithium, type of lithium preparation, and combined treatment with neuroleptics. We found the expected signi­ficant correlation between age and GFR but not between age and maximal osmo­lality.

Of side effects, reported by the patient, (thirst, polyuria, increased frequency) thirst may be an indicator of reduced concentrating ability or increased urine volume.

Some of these negative findings are at variance with several other similar studies (16). This is particularly true of time-on-lithium which in eight out of 10 studies was found to correlate significantly and negatively with maximal osmolality. The reason for the discrepancy is probably to be found in the DDAVP-test. It was not possible to guarantee 12 h absolute water abstinence in this outpatient population. A lack of compliance would tend to weaken a possible relationship between time-on-lithium and maximal osmolality. Another reason to believe that our method may have weakened statistical correlations is the finding of no correlation between age and osmolality.

The number of patients below a defined normal limit for osmolality would also be influenced by patient compliance. Low compliance would increase that number. The lower normal limit was set at 800 mosm/kg H2O, as in other similar studies. Since tubular function, like glomerular function, diminishes with age, the 800 limits can be questioned as being too high and too rigid. Finally, our data suggest that known somatic disorders may have had an influence in the direction of lower­ing maximal osmolality. We, therefore, conclude that our finding of 51 % of the patients below normal on the DDAVP- test gives a too negative picture of the distal tubular function in this population.

Endogenous creatinine clearance is an unreliable measurement of glomerular function, particularly when the patient collects urine at home and only once, as in our study. To compensate for the weakness of the method, we introduced the 51Cr-EDTA-method as a checking device. Preferably, that method should have been used on all the patients. This was not considered feasible at the time in view of the capacity of the laboratory. By checking only those patients who fell below 90 ml/min on creatinine clearance, we introduced a bias in the picture of GFR. This is shown by Table 3, which indicates that endogenous creatinine clear­ance generally renders higher values than the 51Cr-EDTA-method. This finding was confirmed by Wallin et al. (17) who used both methods in 185 patients. They found that the value for creatinine clearance was 23 % higher than that for 51Cr-EDTA- clearance.

In addition, with our method, we were able to correct unusually low and clearly false creatinine clearance values. High false values were only corrected in a few patients while others may have escaped correction. We, therefore, conclude that our finding of four (3 %) patients with GFR slightly below normal gives too positive a picture of the glomerular function.

Apparently, the influence of lithium is more pronounced for tubular than for glomerular function. This is in accordance with most other studies. An extensive dis­cussion of the implications of this study for kidney function and for lithium treat­ment is postponed to a general discussion of studies done on kidney function in lithium patients. Such a discussion appears separately in this issue of Acta Psychiat- Rica Scandinavica (16).

Conclusions from the present study related to introductory questions

1) Lithium does, in some patients, cause a disturbance in kidney function.

2) Concentrating ability is the function mainly affected, while filtrating ability is much less so. Approximately half of our patients had a maximal U-osmolality be­low 800 mosm/kg H2O, four had a mod­erately reduced GFR according to the 51Cr-EDTA-method. Our methods do not allow any conclusions as to the perma­nence of the disturbance. These figures probably mean an underestimate of tubu­lar function, an overestimate of glom­erular function.

3) There is no significant correlation be­tween kidney function and psychiatric diagnosis.

4) There is no significant correlation be­tween kidney function and the following treatment variables: Average serum lithi­um, time-on-lithium, and the total amount of lithium ingested (expressed as the product of the two before-mentioned). Due to methodological sources of error, such a correlation is, however, not excluded.

5) Different lithium preparations do not affect kidney function differently,

6) Treatment with neuroleptics does not affect kidney function.

Cistanche can improve kidney function

Acknowledgments

Multiple, stepwise regression analysis and covariance analysis were performed by Steffan Ekblom and Christer Moller, Statistiska Forskningsgruppen, Uni­versity of Stockholm, using a SPSS Batch System. Dr. Michael Feinberg, M.D., Ph.D., at the Adult Psychiatric Department, University Hospital of Michi­gan, Ann Arbor, Michigan, U.S.A., gave valuable help in translating and editing this paper.

From: ' Kidney function in an unselected lithium population' by H. Bendz, et al

--- Acta Psychiatr. Scand. 1983:68:32%334

References

1. Schou M. Lithium studies. I. Toxicity. Acta Pharmacol Toxicol 1958:75:70-84.

2. Angrist BM, Gershon S, Levitan S J, Blumberg A G. Lithium-induced diabetes insipidus-like syn­drome. Compr Psychiatry 1970:72:141-146.

3. Baldessarini R J, Lipinski J F. Lithium salts: 1970-1975. Ann Intern Med 1975:83:527-533.

4. Johnson F N, Vacaflor L. Lithium side-effects and toxicity: mechanisms. In: Johnson F N, ed. Lithium research and therapy. New York: Aca­demic Press, 1975.

5. Schou M, Amdisen A, Trap-Jensen J. Lithium poisoning. Am J Psychiatry 1968:725:520-527.

6. Lindop G B M, Padfield L. The renal pathology in a case of lithium-induced diabetes insipidus. J Clin Pathol 1975:28:472.

7. Hestbech J, Hansen H E, Amdisen A, Olsen S. Chronic renal lesions following long-term treat­ment with lithium. Kidney Int 1977:12:205-213.

8. Hansen H E, Hestbech J, Olsen St Amdisen A. Renal function and renal pathology in patients with lithium-induced impairment of renal con­centrating ability. Proc Eur Dial Transplant As­soc 1977:74:518-527.

9. Perris C, The Bipolar unipolar dichotomy and the need for consistent terminology. Neuropsy­chobiology 1975:7:65-69.

10. Nurnberger Jr. J, Roose S P, Dunner DL, Fieve R R. Unipolar mania: A distinct clinical entity? Am J Psychiatry 1979:736:1420-1423.

11. Perris C. A study of cycloid psychoses. Acta Psychiatr Scand 1 卯4:Suppl. 253.

12. Bleuler E. Lehrbuch der Psychiatric, 12th ed. Berlin: Springer-Verlag, 1972.

13. Delin K, Aurell M, Ewald J. Urinary concen­tration test with desmopressin. Br Med J 1978: 957:757-758.

14. Brochner-Mortensen J. A simple method for the determination of the glomerular filtration rate. Scand J Clin Lab Invest 1972:30:271-274.

15. Granfirus G, Aurell M. Reference values for slCr- EDTA clearance as a measure of glomerular fil­tration rate. Scand J Clin Lab Invest 1981:42: 611-616.

16. Bendz H. Kidney function in lithium treatment patients: A literature survey. Acta Psychiatr Scand 1983:68:303-324,

17. Wallin L, Alling C, Aurell M. Impairment of renal function in patients on long-term lithium treatment. Clin Nephrol 1982:18:23-28.