Drugs Aging (2014) 31:911–916 DOI 10.1007/s40266-014-0221-1
SHORT COMMUNICATION
Lithium Dosing and Serum Concentrations Across the Age Spectrum: From Early Adulthood to the Tenth Decade of Life Soham Rej • Serge Beaulieu • Marilyn Segal • Nancy C. P. Low • Istvan Mucsi • Christina Holcroft Kenneth Shulman • Karl J. Looper
•
Published online: 21 October 2014 Ó Springer International Publishing Switzerland 2014
Abstract Objectives Little is known about how lithium should be dosed to achieve therapeutic but safe serum concentrations in older adults. In this paper, we investigate how the lithium dose–concentration ratio changes across the lifespan. Methods This was a cross-sectional analysis of 63 current lithium users aged 20–95 years using data from McGLIDICS (the McGill Geriatric Lithium-Induced Diabetes Insipidus Clinical Study). Participants underwent blood and urine tests, including serum lithium concentrations. Multivariate analyses were conducted to evaluate potential correlates of the lithium dose–concentration ratio. Results We found that between the ages of 40–95 years, the total daily dose of lithium required to achieve a given serum concentration decreases threefold (500 vs. 1,500 mg for 1.0 mmol/L). Greater age, once-daily dosing, and lower renal function (estimated glomerular filtration rate) were independently associated with a lower lithium dose–concentration ratio. S. Rej (&) S. Beaulieu M. Segal N. C. P. Low K. J. Looper Department of Psychiatry, McGill University, Irving Ludmer Research and Training Building, 1033 Pine Avenue West, Montreal, QC H3A 1A1, Canada e-mail: [email protected] I. Mucsi Division of Nephrology, Department of Medicine, Royal Victoria Hospital, McGill University, Montreal, QC, Canada C. Holcroft Biostatistician, Boston, MA, USA K. Shulman Division of Geriatric Psychiatry, Department of Psychiatry, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
Conclusions The lithium dose required to achieve a given serum lithium concentration decreases threefold from middle to old age, with this trend continuing into the ninth and tenth decades of life. In order to avoid lithium toxicity in aging patients, continued serum concentration monitoring and judicious dose reduction may be required, particularly in those patients with reduced renal function.
Key Points The daily dose of lithium required to achieve a given serum concentration may decrease up to threefold over the lifespan, particularly between the ages of 40 and 95 years. This change appears to be largely explained by the declining estimated glomerular filtration rate observed in aging lithium users. Judicious lithium concentration and renal monitoring as well as dose adjustment can help prevent lithium toxicity in aging lithium users, who are otherwise at high risk for lithium toxicity.
1 Introduction Lithium continues to be an important treatment in bipolar disorder and unipolar depression [1]. However, the therapeutic window of lithium is relatively narrow. In younger adults, lithium concentrations 0.6–1.0 mmol/L are often necessary for bipolar disorder maintenance, but concentrations[1.2 mmol/L are often toxic [2]. Toxicity is a clinical concern, particularly in older adults [3]: although
912
therapeutic concentrations in geriatric patients may be similar to those in younger adults [4], concentrations \0.8 mmol/L may be necessary to prevent toxicity [5]. Certain experts recommend even lower concentrations, such as 0.5–0.6 mmol/L in geriatric patients, especially when there is co-morbid renal disease, diabetes insipidus, or potential for pharmacokinetic drug interactions (e.g., diuretics) [6]. This issue is complicated by a decrease in the lithium dose–concentration ratio with increasing age [7]. The dose of lithium required to achieve a particular lithium concentration has been estimated to decrease by 35–50 % between the ages of 40 and 65 years [8]. However, there are only five studies examining a range of ages in lithium users [9–13], including a limited number of older adults, with almost no data on patients aged 80 years and older [7, 8]. The primary objective of this study is to describe how the ratio between lithium dose and lithium concentration varies across the lifespan, especially in older adults. In addition, the study examines whether age and other factors (body weight [7, 8], estimated glomerular filtration rate [eGFR] [3, 7], once-daily vs. twice-daily dosing [3, 14]) were independently associated with the lithium dose–concentration ratio.
2 Methods McGLIDICS (the McGill Geriatric Lithium-Induced Diabetes Insipidus Clinical Study) [15] was a cross-sectional study of 100 geriatric (age C65 years) and adult (age 18–64 years) outpatients with current or previous lithium exposure. Between 25 May 2011 and 28 August 2012, 103 consecutive outpatients were recruited at four geriatric psychiatry clinics and two adult mood disorder clinics affiliated with McGill University (Montreal, QC, Canada) and University of Toronto (Toronto, ON, Canada). All patients known to the clinics to have current or past lithium use were approached for recruitment. One adult patient approached for recruitment refused to participate. Three patients withdrew consent. In accordance with the Declaration of Helsinki, each patient provided written consent and ethics approval had been obtained at each of the participating sites. In this study, data on McGLIDICS patients who had both current lithium serum concentrations and doses available were analyzed. Of the 100 patients in the study, 73 were current lithium users, of whom 65 had current lithium concentrations available. Of those with lithium concentrations, 63 patients also had their current lithium dose available. Patients were questioned about their current lithium dose and dosing schedule (once daily vs. twice daily), which was confirmed with their clinicians and their medical charts. As part of the study, patients were also asked to undergo a set of serum and urine tests after 10 h of
S. Rej et al.
fasting and water restriction, including serum lithium concentration, urine osmolality, and eGFR. They were instructed to take their last dose of lithium 12 h prior to their blood test. The main outcome was lithium dose–concentration ratio, which was defined as the total daily dose divided by the serum concentration (lithium dose–concentration ratio = lithium dose/lithium concentration). 2.1 Statistical Analysis Geriatric and adult patients were compared on clinical and demographic variables using Chi-square, Fisher’s exact, Student’s t test, and Mann–Whitney U test, as appropriate. Bivariate correlations were performed between the lithium dose concentration and potential correlates: age [3, 7, 8], body weight [7, 8], eGFR [3, 7], once-daily versus twicedaily dosing [3, 14], concurrent hydrochlorothiazide, loop diuretics, angiotensin II receptor antagonists (blockers) (ARBs)/ACE inhibitors (ACEIs), non-steroidal antiinflammatory drugs (NSAIDs), or aspirin use, using Pearson’s, Spearman’s, Student’s t, and Mann–Whitney U tests, as appropriate. Using variables with significant bivariate correlations, we tested for collinearity and conducted multiple linear regression using lithium dose–concentration ratio as the dependent variable. A two-tailed a of 0.05 was used to determine statistical significance and all analyses were performed using SPSSÒ 20.0 (IBM, Chicago, IL, USA).
3 Results Of the 63 lithium outpatients analyzed, there were 28 geriatric and 35 adult patients ranging from 20 to 95 years of age. Their clinical and demographic characteristics are described in Table 1 and the relationship between age and both prescribed lithium dose and serum concentration is illustrated in Fig. 1a, b. Between the ages of 30 and 95 years, the total daily dose of lithium required to achieve a given serum concentration decreased threefold (538 vs. 1521 mg for 1.0 mmol/L). The dose–concentration ratio appeared to decrease significantly starting from 40 years of age with a marked decrease around 80 years (Fig. 1c; Table 2). On univariate testing, lower lithium dose–concentration ratios were associated with advanced age (rho = -0.65, p \ 0.001), decreased weight (rho = 0.25, p = 0.051), lower eGFR (rho = 0.56, p \ 0.001), and once-daily dosing (U = 141, p = 0.008), but not with sex, urine osmolality, or concurrent hydrochlorothiazide, loop diuretics, ARBs/ACEIs, NSAIDs, or aspirin use [3]. In a multivariate model including age, body weight, eGFR, and once-daily
Lithium Dosing and Serum Concentrations Across the Age Spectrum
913
Table 1 Clinical and demographic characteristics of patients (n = 63) Variable
All patients (n = 63)
Demographics Age (years) [mean (±SD)]
60.0 (±19.8)
% Male [n (%)]
31 (49.2)
Weight (kg) [mean (±SD)]
73.9 (±18.2)
Psychiatric diagnoses [n (%)] Unipolar depression
7 (11.1)
Bipolar disorder
56 (88.9)
Schizoaffective disorder
0 (0)
Lithium use parameters Mean lithium duration (years) [mean (±SD)]
10.9 (±9.71)
Mean lithium duration (years): geriatric patients, age [65 years (n = 28) [mean (±SD)] Mean lithium duration (years): adult patients, age 18–64 years (n = 35) [mean (±SD)]
15.1 (±10.8) 7.5 (±7.3)
Lithium use duration [n (%)] \5 years
13 (20.6)
5–10 years
22 (34.9)
10–20 years
19 (30.2)
C20 years
9 (14.3)
Current lithium dose (mg/day) [mean (±SD)]
693.7 (±389.8)
Current lithium concentration (mmol/L) [mean (±SD)]
0.65 (±0.24)
Daily lithium dose–serum concentrationratio (mg/mmol/L) [mean (±SD)]
1,088.9 (±575.8)
Twice-daily lithium dosing [n (%)]
11 (17.5)
Lithium carbonate (short-acting formulation) [n (%)]
63 (100)
Renal parameters [mean (±SD)] Renal function (eGFR; mL/min/1.73 m2)
79.6 (±23.7)
Urine osmolality (mOsm/Kg)
489.9 (±168.6)
Medications (n = 62) [n (%)] Loop diuretic
1 (1.6)
Hydrochlorothiazide
5 (8.2)
ACEI or ARB
8 (13.1)
NSAIDs or COX-2 inhibitor
1 (1.6)
Aspirin (all 80 mg/day)
5 (8.2)
Any antidepressant
27 (43.5)
Any antipsychotic
31 (50.0)
Valproate
7 (11.3)
Lamotrigine
6 (9.7)
Benzodiazepine
17 (27.4)
ACEI ACE inhibitor, ARB angiotensin II receptor antagonist (blocker), COX-2 cyclo-oxygenase 2, eGFR estimated glomerular filtration rate, NSAIDs non-steroidal anti-inflammatory drugs, SD standard deviation
versus twice-daily dosing, only advanced age (b = -0.37, p = 0.015) and once-daily dosing (b = 0.29, p = 0.008) were independently associated with a lower lithium dose– concentration ratio (Table 3, ‘‘Step 1’’). Since age and eGFR had high collinearity (r = -0.74 [ 0.30), we repeated the model excluding age (Table 3, ‘‘Step 2’’). This time, in addition to once-daily dosing (b = 0.30, p = 0.02), decreased eGFR was also an important predictor of decreased lithium dose–concentration ratio
(b = 0.52, p \ 0.001). Both models explained almost the same amount of the variance in the lithium dose–concentration ratio (R2 = 0.46).
4 Discussion We found that from age 30 to 95 years, the lithium dose required to achieve 1.0 mmol/L decreased threefold (538
914
S. Rej et al.
Daily Lithium Dose (mg)
a 2000 1800 1600 1400 1200 1000 800 600 400 200 0
0
20
40
60
80
100
80
100
Age (years)
Serum Lithium Level (mmol/L)
b
1.4 1.2 1 0.8 0.6 0.4 0.2 0
0
20
40
60
Age (years)
y = -0.0046x + 0.9321 R2 = 0.1338
Lithium Daily Dose-Serum Level Ratio (mg/mmol/L)
c 3500 3000 2500 2000 1500 1000 500 0
0
20
40
60
Age (years)
80
100
y = -15.119x + 1974.8 R2 = 0.3042
Fig. 1 a Lithium dose as a function of age (n = 63). b Serum lithium concentration as a function of age (n = 63). c Ratio of lithium dose– serum lithium concentration as a function of age (n = 63)
vs. 1521 mg/mmol/L). Previous studies had reported a 33 % drop in the lithium dose–serum concentration ratio in the transition between middle-age and late life [7]; however, they examined a narrower age range (e.g. 40–65 years), with smaller geriatric sample sizes. Strengths of our study included: (1) patients ranging from 20 to 95 years of age; (2) a larger sample of patients [65 years than previously examined; (3) specifically analyzing the Table 2 Lithium dose, concentration, and dose– concentration ratio by age category
All data are given as mean ± standard deviation
Age category (years)
Lithium dose (mg)
lithium dose–concentration ratio across the lifespan; and (4) the use of a multivariate approach to identify potential correlates of the lithium dose–concentration ratio. Taken together, our results extend previous work in this area and suggest that the lithium dose–concentration ratio may actually decrease over the lifespan by more than the 25–50 % previously reported [7]. This highlights the need for continued judicious dose titration both in the years preceding old age as well as during late life. From our data, increased age and eGFR were strongly associated with lower lithium dose–serum concentration ratios (the lithium dose needed to achieve a given serum lithium concentration). Given that both age and eGFR were both highly inter-correlated and explained a very similar amount of the variance in lithium dose–concentration ratio, we hypothesize that age lowers the lithium dose–concentration ratio by affecting eGFR. Previous work supports this concept: lithium clearance is often affected proportionally to the degree of renal glomerular function impairment [3, 7]. Additionally, although there can be significant inter-individual variability, eGFR generally decreases at a linear rate of 0.8 mL/min/1.73 m2 [16], starting from 40 years of age. This fits nicely with our data, which also show a relatively linear decline in lithium dose– concentration ratio from 40 to 95 years (Fig. 1). These findings have important clinical implications. Since lithium users, particularly older patients, are often poorly monitored [17], declining lithium dose–concentration ratios put them at risk for a cascade of events: lithium toxicity, medical hospitalization, lithium discontinuation, with high subsequent rates of mood disorder relapse and psychiatric hospitalization [18, 19]. In light of this, regular monitoring of serum lithium concentrations and eGFR in aging lithium patients, especially those with declining renal function, may help prevent lithium toxicity. Also, patients with once-daily lithium dosing tended to require lower doses to achieve a given serum concentration (lower dose–concentration ratios). Although once-daily dosing may be protective against diabetes insipidus [3, 14], we did not find a correlation between urine osmolality and lithium dose–concentration ratio. We suspect that even with equal lithium doses, once-daily dosing leads to higher serum concentrations, primarily due to pharmacokinetic effects [7]. In a recent randomized controlled trial, compared with patients taking lithium twice daily, those Lithium concentration (mmol/L)
Lithium dose–concentration ratio (mg/mmol/L)
20–44 (n = 13)
1,032.7 (±312.5)
0.80 (±0.19)
1,318.1 (±369.4)
45–64 (n = 22)
888.6 (±298.8)
0.68 (±0.26)
1450.7 (±604.0)
65–79 (n = 19)
461.8 (±228.1)
0.59 (±0.20)
800.7 (±351.5)
80–95 (n = 9)
216.7 (±69.6)
0.50 (±0.17)
459.9 (±153.7)
Lithium Dosing and Serum Concentrations Across the Age Spectrum
915
Table 3 Multiple linear regression model with lithium dose–concentration ratio (lithium dose divided by lithium serum concentration) as dependent variable (n = 63) Model
Predictor variables
B (95 % CI)
Step 1
Age
-9.83 (-17.7 to -2.00)
Step 2
Standard error 3.91
b
p value
-0.37
0.015
Body weight (kg)
2.00 (-4.42 to 8.44)
3.21
0.069
0.53
eGFR (mL/min/1.73 m2)
4.63 (-2.01 to 11.3)
3.31
0.21
0.17
0.29
0.008
Twice-daily lithium dosing
411.4 (114.0 to 708.8)
Body weight
5.01 (-2.82 to 12.8)
3.88
0.17
0.20
eGFR
11.4 (5.85 to 16.9)
2.73
0.52
\0.001
Twice-daily lithium dosing
397.9 (64.7 to 731.1)
0.30
0.020
2
148.4
165.1
2
R = 0.46 (Step 1), R = 0.46 (Step 2) Italics indicate variables that are statistically significant in the multiple linear regression models (p \ 0.05) CI confidence interval, eGFR estimated glomerular filtration rate
receiving once-daily dosing had similar improvement in mania and no differences in their neurological toxicity profile but required lower doses to achieve higher lithium concentrations [14]. Furthermore, even though brain lithium concentrations appear to correlate more with 12-h serum lithium concentrations than with dosing schedules, once-daily lithium use is not associated with higher rates of neurological toxicity [20]. Taken together, this suggests that although once-daily lithium use may produce lower lithium dose–concentration ratios, this may not necessarily be harmful for patients.
middle and old age, with this trend continuing into the ninth and tenth decades of life. This is concerning since \30 % of clinicians perform the National Institute for Health and Care Excellence (NICE)-recommended lithium concentration and eGFR monitoring every 3 months in older lithium users [23]. Therefore, in order to avoid lithium toxicity in aging patients, it is imperative that clinicians perform regular serum concentration and eGFR monitoring and judicious dose adjustment. This is particularly important in patients with reduced renal function, who may often have elevated lithium dose–concentration ratios.
4.1 Limitations In comparison with other investigations in this area [8], our sample size was reasonable and included patients up to 95 years old. Nonetheless, the size of our study limited our multivariate analyses. Along similar lines, (thankfully) \13 % of lithium users were also using either a diuretic, ACEI/ARB, or NSAID, making it difficult to exclude the previously well-established effect of these medications on the lithium dose–concentration ratio [3]. Also, 10-h water restriction may have theoretically affected eGFR and lithium concentrations, although the clinical impact of this is unclear. Future longitudinal studies could assess the potential consequences of decreasing lithium dose–concentration ratios on rates of lithium toxicity, medical health service utilization, and mortality in older adults. Given the difficulty of recruiting older lithium patients [21], a meta-regression [22] of existing studies or pharmaco-epidemiological studies using administrative health databases could examine the effects of concurrent medical pharmacotherapy.
5 Conclusion The required lithium dose needed to achieve a given serum lithium concentration appears to decrease threefold in
Acknowledgments We wish to thank Ms. Nicole Bissonnette, Ms. Jean Enright, Dr. Pablo Cervantes, Ms. Sybille Saury, Dr. Rene Desautels, Ms. Rosi Abitbol, Mr. Brian Weixi Li, Ms. Lori Young, Dr. Gershon Frisch, Ms. Dilshad Ratansi, and colleagues at McGill University and University of Toronto. Without their incredible support and encouragement, the implementation of this study would not have been possible. For this study, Soham Rej received Master’s training awards from the Canadian Institutes of Health Research (CIHR) and the Federation de Recherche en Sante Quebec (FRSQ). Conflicts of interest Serge Beaulieu receives research support [CIHR, FRSQ, NARSAD, RSMQ, Stanley Foundation, AstraZeneca, Biovail, Bristol Myers Squibb (BMS), Eli Lilly, Janssen-Ortho, Lundbeck, Merck-Frosst, Novartis, Otsuka, Pfizer, Servier]; speaker bureau fees [AstraZeneca, Biovail, BMS, Eli Lilly, GlaxoSmithKline (GSK), Janssen-Ortho, Lundbeck, Organon, Otsuka, Wyeth Pfizer, Sunovion]; and serves as a consultant/on advisory boards (AstraZeneca, BMS, Eli Lilly, Forest Laboratories, GSK, Janssen-Ortho, Lundbeck, Merck, Pfizer, Otsuka, Servier). Serge Beaulieu did not receive any specific funding for this paper. Soham Rej, Marilyn Segal, Nancy Low, Istvan Mucsi, Christina Holcroft, Kenneth Shulman, and Karl Looper have no conflicts to declare.
References 1. Yatham LN, Kennedy SH, Parikh SV, Schaffer A, Beaulieu S, Alda M, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar
916
2.
3.
4.
5.
6.
7.
8. 9.
10. 11.
12.
13.
S. Rej et al. Disorders (ISBD) collaborative update of CANMAT guidelines for the management of patients with bipolar disorder: update 2013. Bipolar Disord. 2013;15(1):1–44. doi:10.1111/bdi.12025. Hopkins HS, Gelenberg AJ. Serum lithium levels and the outcome of maintenance therapy of bipolar disorder. Bipolar Disord. 2000;2(3 Pt 1):174–9. Rej S, Herrmann N, Shulman K. The effects of lithium on renal function in older adults–a systematic review. J Geriatr Psychiatry Neurol. 2012;25(1):51–61. doi:10.1177/0891988712436690. Al Jurdi RK, Marangell LB, Petersen NJ, Martinez M, Gyulai L, Sajatovic M. Prescription patterns of psychotropic medications in elderly compared with younger participants who achieved a ‘‘recovered’’ status in the systematic treatment enhancement program for bipolar disorder. Am J Geriatr Psychiatry. 2008;16(11):922–33. doi:10.1097/JGP.0b013e318187135f. Rej S, Looper K, Segal M. The effect of serum lithium levels on renal function in geriatric outpatients: a retrospective longitudinal study. Drugs Aging. 2013. doi:10.1007/s40266-013-0068-x. Wijeratne C, Draper B. Reformulation of current recommendations for target serum lithium concentration according to clinical indication, age and physical comorbidity. Aust N Z J Psychiatry. 2011;45(12):1026–32. doi:10.3109/00048674.2011.610296. Sproule BA, Hardy BG, Shulman KI. Differential pharmacokinetics of lithium in elderly patients. Drugs Aging. 2000;16(3): 165–77. Eastham JH, Jeste DV, Young RC. Assessment and treatment of bipolar disorder in the elderly. Drugs Aging. 1998;12(3):205–24. Hewick DS, Newbury P, Hopwood S, Naylor G, Moody J. Age as a factor affecting lithium therapy. Br J Clin Pharmacol. 1977;4(2):201–5. Slater V, Milanes F, Talcott V, Okafor KC (1984) Influence of age on lithium therapy. South Med J 77 (2):153–154, 158. Vestergaard P, Schou M. The effect of age on lithium dosage requirements. Pharmacopsychiatry. 1984;17(6):199–201. doi:10. 1055/s-2007-1017438. Greil W, Stoltzenburg MC, Mairhofer ML, Haag M. Lithium dosage in the elderly. A study with matched age groups. J Affect Disord. 1985;9(1):1–4. Coppen A, Abou-Saleh MT. Lithium therapy: from clinical trials to practical management. Acta Psychiatr Scand. 1988;78(6):754–62.
14. Singh LK, Nizamie SH, Akhtar S, Praharaj SK. Improving tolerability of lithium with a once-daily dosing schedule. Am J Ther. 2011;18(4):288–91. doi:10.1097/MJT.0b013e3181d070c3. 15. Rej S, Segal M, Low NC, Mucsi I, Holcroft C, Shulman K, Looper K. The McGill Geriatric Lithium-Induced Diabetes Insipidus Clinical Study (McGLIDICS). Can J Psychiatry. 2014;59(6):327–34. 16. Weinstein JR, Anderson S. The aging kidney: physiological changes. Adv Chronic Kidney Dis. 2010;17(4):302–7. doi:10. 1053/j.ackd.2010.05.002. 17. Head L, Dening T. Lithium in the over-65s: who is taking it and who is monitoring it? A survey of older adults on lithium in the Cambridge Mental Health Services catchment area. Int J Geriatr Psychiatry. 1998;13(3):164–71. 18. Fahy S, Lawlor BA. Discontinuation of lithium augmentation in an elderly cohort. Int J Geriatr Psychiatry. 2001;16(10):1004–9. 19. Ross J. Discontinuation of lithium augmentation in geriatric patients with unipolar depression: a systematic review. Can J Psychiatry. 2008;53(2):117–20. 20. Jensen HV, Plenge P, Stensgaard A, Mellerup ET, Thomsen C, Aggernaes H, Henriksen O. Twelve-hour brain lithium concentration in lithium maintenance treatment of manic-depressive disorder: daily versus alternate-day dosing schedule. Psychopharmacology (Berl). 1996;124(3):275–8. 21. Young RC, Schulberg HC, Gildengers AG, Sajatovic M, Mulsant BH, Gyulai L, et al. Conceptual and methodological issues in designing a randomized, controlled treatment trial for geriatric bipolar disorder: GERI-BD. Bipolar Disord. 2010;12(1):56–67. doi:10.1111/j.1399-5618.2009.00779.x. 22. Kempton MJ, Geddes JR, Ettinger U, Williams SC, Grasby PM. Meta-analysis, database, and meta-regression of 98 structural imaging studies in bipolar disorder. Arch Gen Psychiatry. 2008;65(9):1017–32. doi:10.1001/archpsyc.65.9.1017. 23. Collins N, Barnes TR, Shingleton-Smith A, Gerrett D, Paton C. Standards of lithium monitoring in mental health trusts in the UK. BMC Psychiatry. 2010;10:80. doi:10.1186/1471-244X-10-80.
SHORT COMMUNICATION
Lithium Dosing and Serum Concentrations Across the Age Spectrum: From Early Adulthood to the Tenth Decade of Life Soham Rej • Serge Beaulieu • Marilyn Segal • Nancy C. P. Low • Istvan Mucsi • Christina Holcroft Kenneth Shulman • Karl J. Looper
•
Published online: 21 October 2014 Ó Springer International Publishing Switzerland 2014
Abstract Objectives Little is known about how lithium should be dosed to achieve therapeutic but safe serum concentrations in older adults. In this paper, we investigate how the lithium dose–concentration ratio changes across the lifespan. Methods This was a cross-sectional analysis of 63 current lithium users aged 20–95 years using data from McGLIDICS (the McGill Geriatric Lithium-Induced Diabetes Insipidus Clinical Study). Participants underwent blood and urine tests, including serum lithium concentrations. Multivariate analyses were conducted to evaluate potential correlates of the lithium dose–concentration ratio. Results We found that between the ages of 40–95 years, the total daily dose of lithium required to achieve a given serum concentration decreases threefold (500 vs. 1,500 mg for 1.0 mmol/L). Greater age, once-daily dosing, and lower renal function (estimated glomerular filtration rate) were independently associated with a lower lithium dose–concentration ratio. S. Rej (&) S. Beaulieu M. Segal N. C. P. Low K. J. Looper Department of Psychiatry, McGill University, Irving Ludmer Research and Training Building, 1033 Pine Avenue West, Montreal, QC H3A 1A1, Canada e-mail: [email protected] I. Mucsi Division of Nephrology, Department of Medicine, Royal Victoria Hospital, McGill University, Montreal, QC, Canada C. Holcroft Biostatistician, Boston, MA, USA K. Shulman Division of Geriatric Psychiatry, Department of Psychiatry, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
Conclusions The lithium dose required to achieve a given serum lithium concentration decreases threefold from middle to old age, with this trend continuing into the ninth and tenth decades of life. In order to avoid lithium toxicity in aging patients, continued serum concentration monitoring and judicious dose reduction may be required, particularly in those patients with reduced renal function.
Key Points The daily dose of lithium required to achieve a given serum concentration may decrease up to threefold over the lifespan, particularly between the ages of 40 and 95 years. This change appears to be largely explained by the declining estimated glomerular filtration rate observed in aging lithium users. Judicious lithium concentration and renal monitoring as well as dose adjustment can help prevent lithium toxicity in aging lithium users, who are otherwise at high risk for lithium toxicity.
1 Introduction Lithium continues to be an important treatment in bipolar disorder and unipolar depression [1]. However, the therapeutic window of lithium is relatively narrow. In younger adults, lithium concentrations 0.6–1.0 mmol/L are often necessary for bipolar disorder maintenance, but concentrations[1.2 mmol/L are often toxic [2]. Toxicity is a clinical concern, particularly in older adults [3]: although
912
therapeutic concentrations in geriatric patients may be similar to those in younger adults [4], concentrations \0.8 mmol/L may be necessary to prevent toxicity [5]. Certain experts recommend even lower concentrations, such as 0.5–0.6 mmol/L in geriatric patients, especially when there is co-morbid renal disease, diabetes insipidus, or potential for pharmacokinetic drug interactions (e.g., diuretics) [6]. This issue is complicated by a decrease in the lithium dose–concentration ratio with increasing age [7]. The dose of lithium required to achieve a particular lithium concentration has been estimated to decrease by 35–50 % between the ages of 40 and 65 years [8]. However, there are only five studies examining a range of ages in lithium users [9–13], including a limited number of older adults, with almost no data on patients aged 80 years and older [7, 8]. The primary objective of this study is to describe how the ratio between lithium dose and lithium concentration varies across the lifespan, especially in older adults. In addition, the study examines whether age and other factors (body weight [7, 8], estimated glomerular filtration rate [eGFR] [3, 7], once-daily vs. twice-daily dosing [3, 14]) were independently associated with the lithium dose–concentration ratio.
2 Methods McGLIDICS (the McGill Geriatric Lithium-Induced Diabetes Insipidus Clinical Study) [15] was a cross-sectional study of 100 geriatric (age C65 years) and adult (age 18–64 years) outpatients with current or previous lithium exposure. Between 25 May 2011 and 28 August 2012, 103 consecutive outpatients were recruited at four geriatric psychiatry clinics and two adult mood disorder clinics affiliated with McGill University (Montreal, QC, Canada) and University of Toronto (Toronto, ON, Canada). All patients known to the clinics to have current or past lithium use were approached for recruitment. One adult patient approached for recruitment refused to participate. Three patients withdrew consent. In accordance with the Declaration of Helsinki, each patient provided written consent and ethics approval had been obtained at each of the participating sites. In this study, data on McGLIDICS patients who had both current lithium serum concentrations and doses available were analyzed. Of the 100 patients in the study, 73 were current lithium users, of whom 65 had current lithium concentrations available. Of those with lithium concentrations, 63 patients also had their current lithium dose available. Patients were questioned about their current lithium dose and dosing schedule (once daily vs. twice daily), which was confirmed with their clinicians and their medical charts. As part of the study, patients were also asked to undergo a set of serum and urine tests after 10 h of
S. Rej et al.
fasting and water restriction, including serum lithium concentration, urine osmolality, and eGFR. They were instructed to take their last dose of lithium 12 h prior to their blood test. The main outcome was lithium dose–concentration ratio, which was defined as the total daily dose divided by the serum concentration (lithium dose–concentration ratio = lithium dose/lithium concentration). 2.1 Statistical Analysis Geriatric and adult patients were compared on clinical and demographic variables using Chi-square, Fisher’s exact, Student’s t test, and Mann–Whitney U test, as appropriate. Bivariate correlations were performed between the lithium dose concentration and potential correlates: age [3, 7, 8], body weight [7, 8], eGFR [3, 7], once-daily versus twicedaily dosing [3, 14], concurrent hydrochlorothiazide, loop diuretics, angiotensin II receptor antagonists (blockers) (ARBs)/ACE inhibitors (ACEIs), non-steroidal antiinflammatory drugs (NSAIDs), or aspirin use, using Pearson’s, Spearman’s, Student’s t, and Mann–Whitney U tests, as appropriate. Using variables with significant bivariate correlations, we tested for collinearity and conducted multiple linear regression using lithium dose–concentration ratio as the dependent variable. A two-tailed a of 0.05 was used to determine statistical significance and all analyses were performed using SPSSÒ 20.0 (IBM, Chicago, IL, USA).
3 Results Of the 63 lithium outpatients analyzed, there were 28 geriatric and 35 adult patients ranging from 20 to 95 years of age. Their clinical and demographic characteristics are described in Table 1 and the relationship between age and both prescribed lithium dose and serum concentration is illustrated in Fig. 1a, b. Between the ages of 30 and 95 years, the total daily dose of lithium required to achieve a given serum concentration decreased threefold (538 vs. 1521 mg for 1.0 mmol/L). The dose–concentration ratio appeared to decrease significantly starting from 40 years of age with a marked decrease around 80 years (Fig. 1c; Table 2). On univariate testing, lower lithium dose–concentration ratios were associated with advanced age (rho = -0.65, p \ 0.001), decreased weight (rho = 0.25, p = 0.051), lower eGFR (rho = 0.56, p \ 0.001), and once-daily dosing (U = 141, p = 0.008), but not with sex, urine osmolality, or concurrent hydrochlorothiazide, loop diuretics, ARBs/ACEIs, NSAIDs, or aspirin use [3]. In a multivariate model including age, body weight, eGFR, and once-daily
Lithium Dosing and Serum Concentrations Across the Age Spectrum
913
Table 1 Clinical and demographic characteristics of patients (n = 63) Variable
All patients (n = 63)
Demographics Age (years) [mean (±SD)]
60.0 (±19.8)
% Male [n (%)]
31 (49.2)
Weight (kg) [mean (±SD)]
73.9 (±18.2)
Psychiatric diagnoses [n (%)] Unipolar depression
7 (11.1)
Bipolar disorder
56 (88.9)
Schizoaffective disorder
0 (0)
Lithium use parameters Mean lithium duration (years) [mean (±SD)]
10.9 (±9.71)
Mean lithium duration (years): geriatric patients, age [65 years (n = 28) [mean (±SD)] Mean lithium duration (years): adult patients, age 18–64 years (n = 35) [mean (±SD)]
15.1 (±10.8) 7.5 (±7.3)
Lithium use duration [n (%)] \5 years
13 (20.6)
5–10 years
22 (34.9)
10–20 years
19 (30.2)
C20 years
9 (14.3)
Current lithium dose (mg/day) [mean (±SD)]
693.7 (±389.8)
Current lithium concentration (mmol/L) [mean (±SD)]
0.65 (±0.24)
Daily lithium dose–serum concentrationratio (mg/mmol/L) [mean (±SD)]
1,088.9 (±575.8)
Twice-daily lithium dosing [n (%)]
11 (17.5)
Lithium carbonate (short-acting formulation) [n (%)]
63 (100)
Renal parameters [mean (±SD)] Renal function (eGFR; mL/min/1.73 m2)
79.6 (±23.7)
Urine osmolality (mOsm/Kg)
489.9 (±168.6)
Medications (n = 62) [n (%)] Loop diuretic
1 (1.6)
Hydrochlorothiazide
5 (8.2)
ACEI or ARB
8 (13.1)
NSAIDs or COX-2 inhibitor
1 (1.6)
Aspirin (all 80 mg/day)
5 (8.2)
Any antidepressant
27 (43.5)
Any antipsychotic
31 (50.0)
Valproate
7 (11.3)
Lamotrigine
6 (9.7)
Benzodiazepine
17 (27.4)
ACEI ACE inhibitor, ARB angiotensin II receptor antagonist (blocker), COX-2 cyclo-oxygenase 2, eGFR estimated glomerular filtration rate, NSAIDs non-steroidal anti-inflammatory drugs, SD standard deviation
versus twice-daily dosing, only advanced age (b = -0.37, p = 0.015) and once-daily dosing (b = 0.29, p = 0.008) were independently associated with a lower lithium dose– concentration ratio (Table 3, ‘‘Step 1’’). Since age and eGFR had high collinearity (r = -0.74 [ 0.30), we repeated the model excluding age (Table 3, ‘‘Step 2’’). This time, in addition to once-daily dosing (b = 0.30, p = 0.02), decreased eGFR was also an important predictor of decreased lithium dose–concentration ratio
(b = 0.52, p \ 0.001). Both models explained almost the same amount of the variance in the lithium dose–concentration ratio (R2 = 0.46).
4 Discussion We found that from age 30 to 95 years, the lithium dose required to achieve 1.0 mmol/L decreased threefold (538
914
S. Rej et al.
Daily Lithium Dose (mg)
a 2000 1800 1600 1400 1200 1000 800 600 400 200 0
0
20
40
60
80
100
80
100
Age (years)
Serum Lithium Level (mmol/L)
b
1.4 1.2 1 0.8 0.6 0.4 0.2 0
0
20
40
60
Age (years)
y = -0.0046x + 0.9321 R2 = 0.1338
Lithium Daily Dose-Serum Level Ratio (mg/mmol/L)
c 3500 3000 2500 2000 1500 1000 500 0
0
20
40
60
Age (years)
80
100
y = -15.119x + 1974.8 R2 = 0.3042
Fig. 1 a Lithium dose as a function of age (n = 63). b Serum lithium concentration as a function of age (n = 63). c Ratio of lithium dose– serum lithium concentration as a function of age (n = 63)
vs. 1521 mg/mmol/L). Previous studies had reported a 33 % drop in the lithium dose–serum concentration ratio in the transition between middle-age and late life [7]; however, they examined a narrower age range (e.g. 40–65 years), with smaller geriatric sample sizes. Strengths of our study included: (1) patients ranging from 20 to 95 years of age; (2) a larger sample of patients [65 years than previously examined; (3) specifically analyzing the Table 2 Lithium dose, concentration, and dose– concentration ratio by age category
All data are given as mean ± standard deviation
Age category (years)
Lithium dose (mg)
lithium dose–concentration ratio across the lifespan; and (4) the use of a multivariate approach to identify potential correlates of the lithium dose–concentration ratio. Taken together, our results extend previous work in this area and suggest that the lithium dose–concentration ratio may actually decrease over the lifespan by more than the 25–50 % previously reported [7]. This highlights the need for continued judicious dose titration both in the years preceding old age as well as during late life. From our data, increased age and eGFR were strongly associated with lower lithium dose–serum concentration ratios (the lithium dose needed to achieve a given serum lithium concentration). Given that both age and eGFR were both highly inter-correlated and explained a very similar amount of the variance in lithium dose–concentration ratio, we hypothesize that age lowers the lithium dose–concentration ratio by affecting eGFR. Previous work supports this concept: lithium clearance is often affected proportionally to the degree of renal glomerular function impairment [3, 7]. Additionally, although there can be significant inter-individual variability, eGFR generally decreases at a linear rate of 0.8 mL/min/1.73 m2 [16], starting from 40 years of age. This fits nicely with our data, which also show a relatively linear decline in lithium dose– concentration ratio from 40 to 95 years (Fig. 1). These findings have important clinical implications. Since lithium users, particularly older patients, are often poorly monitored [17], declining lithium dose–concentration ratios put them at risk for a cascade of events: lithium toxicity, medical hospitalization, lithium discontinuation, with high subsequent rates of mood disorder relapse and psychiatric hospitalization [18, 19]. In light of this, regular monitoring of serum lithium concentrations and eGFR in aging lithium patients, especially those with declining renal function, may help prevent lithium toxicity. Also, patients with once-daily lithium dosing tended to require lower doses to achieve a given serum concentration (lower dose–concentration ratios). Although once-daily dosing may be protective against diabetes insipidus [3, 14], we did not find a correlation between urine osmolality and lithium dose–concentration ratio. We suspect that even with equal lithium doses, once-daily dosing leads to higher serum concentrations, primarily due to pharmacokinetic effects [7]. In a recent randomized controlled trial, compared with patients taking lithium twice daily, those Lithium concentration (mmol/L)
Lithium dose–concentration ratio (mg/mmol/L)
20–44 (n = 13)
1,032.7 (±312.5)
0.80 (±0.19)
1,318.1 (±369.4)
45–64 (n = 22)
888.6 (±298.8)
0.68 (±0.26)
1450.7 (±604.0)
65–79 (n = 19)
461.8 (±228.1)
0.59 (±0.20)
800.7 (±351.5)
80–95 (n = 9)
216.7 (±69.6)
0.50 (±0.17)
459.9 (±153.7)
Lithium Dosing and Serum Concentrations Across the Age Spectrum
915
Table 3 Multiple linear regression model with lithium dose–concentration ratio (lithium dose divided by lithium serum concentration) as dependent variable (n = 63) Model
Predictor variables
B (95 % CI)
Step 1
Age
-9.83 (-17.7 to -2.00)
Step 2
Standard error 3.91
b
p value
-0.37
0.015
Body weight (kg)
2.00 (-4.42 to 8.44)
3.21
0.069
0.53
eGFR (mL/min/1.73 m2)
4.63 (-2.01 to 11.3)
3.31
0.21
0.17
0.29
0.008
Twice-daily lithium dosing
411.4 (114.0 to 708.8)
Body weight
5.01 (-2.82 to 12.8)
3.88
0.17
0.20
eGFR
11.4 (5.85 to 16.9)
2.73
0.52
\0.001
Twice-daily lithium dosing
397.9 (64.7 to 731.1)
0.30
0.020
2
148.4
165.1
2
R = 0.46 (Step 1), R = 0.46 (Step 2) Italics indicate variables that are statistically significant in the multiple linear regression models (p \ 0.05) CI confidence interval, eGFR estimated glomerular filtration rate
receiving once-daily dosing had similar improvement in mania and no differences in their neurological toxicity profile but required lower doses to achieve higher lithium concentrations [14]. Furthermore, even though brain lithium concentrations appear to correlate more with 12-h serum lithium concentrations than with dosing schedules, once-daily lithium use is not associated with higher rates of neurological toxicity [20]. Taken together, this suggests that although once-daily lithium use may produce lower lithium dose–concentration ratios, this may not necessarily be harmful for patients.
middle and old age, with this trend continuing into the ninth and tenth decades of life. This is concerning since \30 % of clinicians perform the National Institute for Health and Care Excellence (NICE)-recommended lithium concentration and eGFR monitoring every 3 months in older lithium users [23]. Therefore, in order to avoid lithium toxicity in aging patients, it is imperative that clinicians perform regular serum concentration and eGFR monitoring and judicious dose adjustment. This is particularly important in patients with reduced renal function, who may often have elevated lithium dose–concentration ratios.
4.1 Limitations In comparison with other investigations in this area [8], our sample size was reasonable and included patients up to 95 years old. Nonetheless, the size of our study limited our multivariate analyses. Along similar lines, (thankfully) \13 % of lithium users were also using either a diuretic, ACEI/ARB, or NSAID, making it difficult to exclude the previously well-established effect of these medications on the lithium dose–concentration ratio [3]. Also, 10-h water restriction may have theoretically affected eGFR and lithium concentrations, although the clinical impact of this is unclear. Future longitudinal studies could assess the potential consequences of decreasing lithium dose–concentration ratios on rates of lithium toxicity, medical health service utilization, and mortality in older adults. Given the difficulty of recruiting older lithium patients [21], a meta-regression [22] of existing studies or pharmaco-epidemiological studies using administrative health databases could examine the effects of concurrent medical pharmacotherapy.
5 Conclusion The required lithium dose needed to achieve a given serum lithium concentration appears to decrease threefold in
Acknowledgments We wish to thank Ms. Nicole Bissonnette, Ms. Jean Enright, Dr. Pablo Cervantes, Ms. Sybille Saury, Dr. Rene Desautels, Ms. Rosi Abitbol, Mr. Brian Weixi Li, Ms. Lori Young, Dr. Gershon Frisch, Ms. Dilshad Ratansi, and colleagues at McGill University and University of Toronto. Without their incredible support and encouragement, the implementation of this study would not have been possible. For this study, Soham Rej received Master’s training awards from the Canadian Institutes of Health Research (CIHR) and the Federation de Recherche en Sante Quebec (FRSQ). Conflicts of interest Serge Beaulieu receives research support [CIHR, FRSQ, NARSAD, RSMQ, Stanley Foundation, AstraZeneca, Biovail, Bristol Myers Squibb (BMS), Eli Lilly, Janssen-Ortho, Lundbeck, Merck-Frosst, Novartis, Otsuka, Pfizer, Servier]; speaker bureau fees [AstraZeneca, Biovail, BMS, Eli Lilly, GlaxoSmithKline (GSK), Janssen-Ortho, Lundbeck, Organon, Otsuka, Wyeth Pfizer, Sunovion]; and serves as a consultant/on advisory boards (AstraZeneca, BMS, Eli Lilly, Forest Laboratories, GSK, Janssen-Ortho, Lundbeck, Merck, Pfizer, Otsuka, Servier). Serge Beaulieu did not receive any specific funding for this paper. Soham Rej, Marilyn Segal, Nancy Low, Istvan Mucsi, Christina Holcroft, Kenneth Shulman, and Karl Looper have no conflicts to declare.
References 1. Yatham LN, Kennedy SH, Parikh SV, Schaffer A, Beaulieu S, Alda M, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar
916
2.
3.
4.
5.
6.
7.
8. 9.
10. 11.
12.
13.
S. Rej et al. Disorders (ISBD) collaborative update of CANMAT guidelines for the management of patients with bipolar disorder: update 2013. Bipolar Disord. 2013;15(1):1–44. doi:10.1111/bdi.12025. Hopkins HS, Gelenberg AJ. Serum lithium levels and the outcome of maintenance therapy of bipolar disorder. Bipolar Disord. 2000;2(3 Pt 1):174–9. Rej S, Herrmann N, Shulman K. The effects of lithium on renal function in older adults–a systematic review. J Geriatr Psychiatry Neurol. 2012;25(1):51–61. doi:10.1177/0891988712436690. Al Jurdi RK, Marangell LB, Petersen NJ, Martinez M, Gyulai L, Sajatovic M. Prescription patterns of psychotropic medications in elderly compared with younger participants who achieved a ‘‘recovered’’ status in the systematic treatment enhancement program for bipolar disorder. Am J Geriatr Psychiatry. 2008;16(11):922–33. doi:10.1097/JGP.0b013e318187135f. Rej S, Looper K, Segal M. The effect of serum lithium levels on renal function in geriatric outpatients: a retrospective longitudinal study. Drugs Aging. 2013. doi:10.1007/s40266-013-0068-x. Wijeratne C, Draper B. Reformulation of current recommendations for target serum lithium concentration according to clinical indication, age and physical comorbidity. Aust N Z J Psychiatry. 2011;45(12):1026–32. doi:10.3109/00048674.2011.610296. Sproule BA, Hardy BG, Shulman KI. Differential pharmacokinetics of lithium in elderly patients. Drugs Aging. 2000;16(3): 165–77. Eastham JH, Jeste DV, Young RC. Assessment and treatment of bipolar disorder in the elderly. Drugs Aging. 1998;12(3):205–24. Hewick DS, Newbury P, Hopwood S, Naylor G, Moody J. Age as a factor affecting lithium therapy. Br J Clin Pharmacol. 1977;4(2):201–5. Slater V, Milanes F, Talcott V, Okafor KC (1984) Influence of age on lithium therapy. South Med J 77 (2):153–154, 158. Vestergaard P, Schou M. The effect of age on lithium dosage requirements. Pharmacopsychiatry. 1984;17(6):199–201. doi:10. 1055/s-2007-1017438. Greil W, Stoltzenburg MC, Mairhofer ML, Haag M. Lithium dosage in the elderly. A study with matched age groups. J Affect Disord. 1985;9(1):1–4. Coppen A, Abou-Saleh MT. Lithium therapy: from clinical trials to practical management. Acta Psychiatr Scand. 1988;78(6):754–62.
14. Singh LK, Nizamie SH, Akhtar S, Praharaj SK. Improving tolerability of lithium with a once-daily dosing schedule. Am J Ther. 2011;18(4):288–91. doi:10.1097/MJT.0b013e3181d070c3. 15. Rej S, Segal M, Low NC, Mucsi I, Holcroft C, Shulman K, Looper K. The McGill Geriatric Lithium-Induced Diabetes Insipidus Clinical Study (McGLIDICS). Can J Psychiatry. 2014;59(6):327–34. 16. Weinstein JR, Anderson S. The aging kidney: physiological changes. Adv Chronic Kidney Dis. 2010;17(4):302–7. doi:10. 1053/j.ackd.2010.05.002. 17. Head L, Dening T. Lithium in the over-65s: who is taking it and who is monitoring it? A survey of older adults on lithium in the Cambridge Mental Health Services catchment area. Int J Geriatr Psychiatry. 1998;13(3):164–71. 18. Fahy S, Lawlor BA. Discontinuation of lithium augmentation in an elderly cohort. Int J Geriatr Psychiatry. 2001;16(10):1004–9. 19. Ross J. Discontinuation of lithium augmentation in geriatric patients with unipolar depression: a systematic review. Can J Psychiatry. 2008;53(2):117–20. 20. Jensen HV, Plenge P, Stensgaard A, Mellerup ET, Thomsen C, Aggernaes H, Henriksen O. Twelve-hour brain lithium concentration in lithium maintenance treatment of manic-depressive disorder: daily versus alternate-day dosing schedule. Psychopharmacology (Berl). 1996;124(3):275–8. 21. Young RC, Schulberg HC, Gildengers AG, Sajatovic M, Mulsant BH, Gyulai L, et al. Conceptual and methodological issues in designing a randomized, controlled treatment trial for geriatric bipolar disorder: GERI-BD. Bipolar Disord. 2010;12(1):56–67. doi:10.1111/j.1399-5618.2009.00779.x. 22. Kempton MJ, Geddes JR, Ettinger U, Williams SC, Grasby PM. Meta-analysis, database, and meta-regression of 98 structural imaging studies in bipolar disorder. Arch Gen Psychiatry. 2008;65(9):1017–32. doi:10.1001/archpsyc.65.9.1017. 23. Collins N, Barnes TR, Shingleton-Smith A, Gerrett D, Paton C. Standards of lithium monitoring in mental health trusts in the UK. BMC Psychiatry. 2010;10:80. doi:10.1186/1471-244X-10-80.
