jama_241_11_025 (1)

Serum Lithium Levels andLong-term Diuretic UseJames W. Jefferson, MD, Ned H. Kalin, MD

\s=b\In normal volunteers whose conditions were stabilized with lithiumcarbonate, there were no significant changes in serum lithium levels during atwo-week period while they received 40 mg/day of furosemide. Duringanother two-week period, a 50 mg/day dosage of hydrochlorothiazide didcause a significant rise in serum lithium levels. These observations suggestthat a substantial amount of lithium may be reabsorbed in the loop of Henleand that the combined use of lithium and a loop diuretic may not requiremodification of the lithium dosage.

(JAMA 241:1134-1136, 1979)

IT IS widely held that lithium anddiuretics interact adversely, leadingto diuretic-induced lithium retention,an increase in the serum lithiumlevel, and possible lithium intoxication. This interaction has been welldocumented with the thiazides bothin animals' and in man," but littlework has been done with the otherclasses of diuretics.5

When the potassium-sparing diuretic spironolactone was given tofour subjects during a period ofseveral days, there was a consistentincrease in the serum lithium leveldespite little or no change in externallithium balance.6 Experiments withthe loop diuretics (furosemide andethacrynic acid) have been limited tosingle-dose studies, which have shownlithium clearance to be either in-

creased' or unchanged.8 A 72-year-oldwoman did have lithium intoxicationduring treatment with furosemide,but dietary sodium had also beenrestricted, and the role of the drugalone could not be determined.9

Recently, Steele et al10 found thatthe loop of Henle may be a site ofsubstantial lithium reabsorption.This led us to speculate that a drug-blocking reabsorption at this sitemight have a different effect on lithium clearance than the more distallyacting thiazides. The possibility thatthe prolonged use of a loop diureticwould not cause a clinically importantalteration in the serum lithium leveland that the two drugs could be usedtogether without risking lithium retention and toxic effects or necessitating a readjustment of lithiumdosage led to our study.

METHODSSix paid volunteers (five men and one

woman; age, 24 to 40 years) participated inthis open study. They were in good general

health, took no other medications, andgave written informed consent.

Procedure.Each subject took lithiumcarbonate, 300 mg three times a day, for 44days. Their conditions were initially stabilized while they received lithium alone fornine days; they then took lithium plusfurosemide (40 mg/day) for two weeks(days 10 through 23), followed by lithiumalone for seven days, and finally lithiumplus hydrochlorothiazide (50 mg/day) foran additional two weeks (days 31 through44).

Serum lithium determinations were performed by atomic absorption spectropho-tometry on days 6, 8, 10, 13, 15, 17, 20, 22,24, 29, 31, 34, 36, 38, 41, 43, and 45 onsamples of venous blood drawn 12 hoursafter the last dose of lithium.

Subjects continued their usual dailyactivities (all were medical professionals)with no restriction of either diet or exercise. They remained in close contact withthe investigators (who also participated assubjects) and were frequently checked foradverse reactions.

Data Analysis.The study was dividedinto six periods (lithium only, first baseline; lithium and furosemide, first week;lithium and furosemide, second week; lithium only, second baseline; lithium andhydrochlorothiazide, first week; and lithium and hydrochlorothiazide, secondweek). The data (serum lithium levels)were subjected to a repeated measuresanalysis of variance. The Scheffe test(critical P value, .05) was used to comparethe mean serum lithium levels for eachtrial period with each other (the two baseline periods vs the two weeks of lithium-furosemide vs the two weeks of lithium-hydrochlorothiazide).

From the Department of Psychiatry, Universityof Wisconsin Clinical Sciences Center, Madison.

Reprint requests to Department of Psychiatry,University of Wisconsin Clinical Sciences Center,600 Highland Ave, Madison, WI 53792 (Dr Jeffer-son).

Downloaded From: http://jama.jamanetwork.com/ by a Icahn School of Medicine and Mount Sinai Medical Center User on 05/12/2015

RESULTSFive subjects completed the study.

The sixth withdrew on day 15 becauseof a variety of symptoms (Table) thathad begun on day 6 (lithium alone;serum level, 0.44 mEq/L) and wereintensified on days 13 through 15(lithium-furosemide; serum levels,0.72 and 0.71 mEq/L).

Data were pooled from the fivesubjects who completed the study.When group mean serum lithiumlevels were compared for each of sixstudy periods, the following resultswere obtained: (1) no significantdifference between the two lithium-

only baseline periods, (2) no significant difference between the firstweek receiving a diuretic and thesecond week receiving the samediuretic, (3) no significant differencebetween the two lithium-only baseline and the lithium-furosemide periods, (4) significant increase in serumlithium levels during the lithium-hydrochlorothiazide periods as compared with the two lithium-only baseline periods (P

on the medullary and cortical thickascending limb of Henle's loop, wespeculate that a balance is established between the decreased lithiumreabsorption at this site and thecompensatory increase in proximaltubular lithium reabsorption.

Our finding that hydrochlorothiazide did cause a significant (P