European Heart Journal (1992) 13, {Supplement G), 96-103
Adverse reactions to diuretics B. N. C. PRICHARD, C. W. I. OWENS AND A. S. WOOLF
Division of Clinical Pharmacology, University College, London, U.K. KEY WORDS: Thiazides, adverse reactions, hyponatraemia, hypokalaemia, hypercalcaemia, hyperlipidaemia, interactions, potassium retaining diuretics. Diuretics can result in various undesired biochemical changes, such as impotence, skin rashes, nausea, dizziness and lethargy as well as subjective side effects. The side effects are mostly predictable, their effects depending on both the circulatory blood volume and on the transport of water and solute in the renal tubules. Two of the commonest side effects are mild hypovolaemia, when any diuretic is used, and mild hypokalaemia when the non-potassium-sparing diuretics, such as thiazides and frusemide are used. Its occurrence is dose dependent and can be corrected by potassium supplements, but potassium-retaining diuretics, which also correct the often associated fall in serum magnesium, are preferable. Many reports link hypokalaemia with cardiac arrhythmias, but some dispute this association in the absence of the concomitant use ofdigoxin. Hyponatraemia rarely occurs, but can be life threatening. Calcium excretion is markedly reduced, but unlike other electrolyte disturbancesfrom diuretics, this may be valuable: some suggest diuretics have an anti-osteoporotic action. Diuretics increase glucose and insulin resistance and should be used sparingly in diabetics. They rarely cause a non-ketotic hyperosmolar coma. Urate is raised, but clinical gout is not common. Cholesterol elevation has been reported in some studies, but long-term studies indicate that lipid changes are minor. Other rare side effects are not predictable from their pharmacological actions and these include the occurrence of skin rashes, thrombocytopenia, pancreatitis and interstitial nephritis; and ototoxicity from frusemide. Downloaded from by guest on October 25, 2014
Introduction Diuretics have been used in the treatment of hypertension for many years. They possess a wide side effect profile, ranging from subjective effects and idiosyncratic responses to a wide spectrum of metabolic effects. There are also some clinically important interactions with other drugs. Subjective symptoms The MRC1'1 trial in hypertension is the most comprehensive placebo-controlled study, and demonstrated that the most notable side effect from bendrofluazide was the incidence of impotence, which resulted in 19-58 withdrawals from the drug per 1000 patient years. This finding had not hitherto been widely recognised. However, diuretics are not the only antihypertensive drugs associated with impotence; there is a similar incidence with methyldopa and reserpine'21. The patient withdrawals for men with gout were 12-23, for impaired glucose tolerance 9-38 withdrawals per 1000 patient years, for lethargy 6-93, nausea, dizziness or headache 8-56, and constipation 1 -63, all significantly different from placebo. The withdrawals reaching significance for women were, glucose intolerance 601, lethargy 212, nausea, dizziness or headache 16-27, and constipation 212 per 1000 patient years. The MRC1'1 study used what now would be regarded as a high dose of bendofluazide 10 mg. Carlsen et alP] reported a placebo-controlled study of bendrofluazide using l-25mg, 2-5, 50 and lOmg per day with 50-52 patients in each group. The incidence of subjective adverse effects with dosages up to 5 mg were very similar to Correspondence: Prof. B. N. C. Prichard, Division of Clinical Pharmacology, University College London, 5 University Street, London WCIE 6JJ, U.K. 0I95-668X/92/0GOO96 + 08 $08.00/0
1-25
2-5
5-0
10-0
1
Bendrofluazide (mg.day" ) Figure I Effect of 10 weeks administration of graded doses of bendrofluazide on serum creatinine. Placebo n = 52, l-2mg n = 50, 2-5 mg n = 52, 5 0 mg n = 52, 10 mg n = 51. Double-blind parallel group study. (Drawn from data, with permission131). P value, with respect to bendrofluazide 0 mg.
placebo, but moving to 10 mg doubled the total incidence of side effects from 12 out of 52 patients to 24 out of 51 patients. Lowe et al.[A\ in their surveillance study with frusemide, also found a dose-dependent increase in incidence of subjective side effects. Hypovolaemia Hypovolaemia, in its mildest form, is indicated by the small rise in creatinine or urea associated with diuretic administration (Fig. 1)[3], but significant hypovolaemia is much more likely to occur with a loop diuretic. Lowe el a/.'4' found that the incidence of hypovolaemia with frusemide was dose dependent, increasing from 7-5% (n = 347) with doses of 40 mg or less, 17-5% (n = 160) at © 1992 The European Society of Cardiology
Adverse reactions to diuretics 97
Table I
Metabolic side effects of diuretics
1. Hyponatraemia 2. Hypokalaemia 3. Hypomagnesaemia 4. Hypercalcaemia 5. Hyperuricaemia 6. Hyperglycaemia 7. Disturbance lipid metabolism
of commencing thiazide treatment. Besides measures to ensure repletion of sodium in hyponatraemia there is some evidence that magnesium may help to rectify this electrolyte imbalance181. Hypokalaemia and hypomagnesaemia
Downloaded from by guest on October 25, 2014
Hypokalaemia from diuretic administration results from, (i) increased delivery of sodium to the distal tubule so more is available for exchange with potassium (ii) secondary hyperaldosteronism due to volume depletion dosages between 40 and 80 mg, while above 80 mg the and renin secretion and (iii) a shift of potassium into the incidence was 39-7% (n = 78). Hypovolaemia produced cells because of diuretic-induced alkalosis, which results by diuretics may cause a major fall in glomerular fil- from increased bicarbonate resorption by the proximal tration rate (GFR) when used with other drugs such as tubule secondary to hypovolaemia. non-steroidal anti-flammatory agents and angiotensin Carlsen et alP\ in their dose response study of bendroconverting enzyme inhibitors, and this is discussed below. fluazide, found that a near maximum anti-hypertensive effect was observed from l-25mg daily, with similar responses from 2-5, 50 and 10 mg (Fig. 2), whereas there Metabolic side effects was a dose-related fall in potassium levels from 2-5 mg up Diuretics may cause varied metabolic side effects to 10 mg of bendrofluazide (Fig. 3). Similarly, Owens (Table 1). et alP\ in an eight increment dose-response study of xipamide, found the maximum fall in blood pressure was seen with 10 to 20 mg, whereas a fall in potassium only HYPONATRAEMIA Hyponatraemia occurs by a variety of mechanisms as a occurred at 40 mg. The occurrence of hypokalaemia is influenced by result of impairment of free water clearance. The mechanisms include hypovolaemia-induced fall in GFR and potassium intake. Prichard and Brogden'101 reported on enhanced proximal tubule re-absorption: both result in antihypertensive studies of xipamide. If these are the delivery of less fluid to the diluting segment, where arranged according to geography, there is a difference some diuretics, such as thiazides, have the additional between those studies performed in Northern European effect of inhibitory sodium re-absorption. In addition, countries and those where their calorie intake is likely to hypovolaemia will cause both the non-osmotic release of be obtained in a greater proportion from potassium-rich antidiuretic hormone (ADH) and the onset of thirst which fruit and vegetables. While plasma potassium levels can be increased by potassium supplements, Ramsey and his results in water intake. diuretics, Diuretic-induced hyponatraemia is unusual in patients colleagues have shown that potassium-retaining 1 1 11 such as triamterene and spironolactone " , are more under the age of 65 years. Three years into the MRC ' trial as is amiloride'12'131. Additionally, Schnaper there were no important changes in sodium levels in either efficient, 141 men or women, although in both sexes levels compared to et a/.' showed that while potassium-retaining diuretics base line were significantly different from placebo. Small correct potassium, unlike potassium supplements, they increases of 0-9 and 0-6mmol. I"1 were seen in men and also correct magnesium deficiency commonly associated with hypokalaemia (Fig. 4). women respectively. Even common arrhythmias, such as ventricular ectopic There is, however, an important incidence of hyponatraemia in the elderly. Plante and Dessuralt'3' found that beats, are an unfavourable prognostic sign as they are there was an average fall of 141 to 134mmol . 1 " ' from the associated with an increase in the risk of sudden death by administration of 50 mg per day of hydrochlorothiazide a factor of 2-2'131. While some investigations have found for 48 weeks to a group of 19 patients between 65 to 91 years an association between hypokalaemia and cardiac old. Sunderam and Mankikar161 reported 11-3% of hypo- arrhythmias'161, others have not'17181. However, Storstein'151 natraemia incidence (< 130mmol. I"1) in 683 geriatric has shown that some studies indicate an association patients admitted to hospital over a 10 month period, while between hypokalaemia from diuretics and arrhythmias. in 4-5% of the patients the level was under 125mmol .I" 1 . Papdemetriou'181 and his colleagues have found that This was associated with symptoms in these 77 patients of correction of diuretic-induced hypokalaemia from 2-8 to weakness in 28, confusion in 21, postural giddiness in 20, 3-7 mEq . 1 " ' , the administration of diuretics to patients postural hypotension in 17, falls in 13, fits in two, and with left ventricular hypertrophy, even if severe, or transient hemiparesis in one. I n the 49 cases where diuretics diuretics to patients with exercise arrhythmias, had no were responsible for the hyponatraemia, just under a effect on the incidence of arrhythmias. quarter of those were receiving frusemide, as was the case with cyclopenthiazide, while 20% received hydrochlorothiazide amiloride combination. Finally, Ashraf and Hypercalcaemia colleagues'71 described seven patients (Table 2) who Any diuretic may enhance proximal tubular resorption developed dangerous hyponatraemia within a few days of calcium secondary to volume depletion and, in addition,
98 B. N.C.Prichardetal. Table 2
Severe hyponalraemia and thiazides
Age/Sex/In
Na/K
Drug (mg)
Days R.
Time
Outcome
72FBP 50FBP
100/3-1 103/20
10 3
30 h 44h
Paraparesis Coma, died 5/12
75FBP 62MBP
105/1-6 109/2-2
Paralysis 1 Leg Recovered
116/2-9 104/3-0 98/2-4
16 3 7 9 5 5
18h 18 h
68MBP 82FMI 69FBP
HCTZ50 HCTZ50 Metol. 7 HCTZ50 HCTZ50 Metol. 5 HCTZ100 HCTZ25 POLYT.3
14h 48h 56 h
Recovered Coma, died Coma, died 11/52
In = indication; BP = blood pressure; MI = myocardial infarction; Na/K = sodium/potassium levels at presentation in E q . L" 1 ; HCTZ = hydrochlorothiazide; Metol = metolozone; POLYT = polythiazide; Days R = days of treatment before presentation; Time = time in hours required to raise serum sodium to 130 mEq . L~'orabove. After Ashraf etal.m.
Figure 2 Effect of graded doses of bendrofluazide on blood pressure control (for details see Fig. 1). (Drawn from data, with permissionpl)H = week 4; 0 = mean week 10-12.
Hyperuricaemia
An elevation of urate from thiazides is due to reduced blood volume, increased re-absorption by the proximal tubule and to reduced uric acid excretion in the proximal tubule as a result of competition by the diuretic for the organic acid secreting pathway. Carlsen et a/.'31 demonstrated a dose-dependent increase in urate levels from 1 -25 to 10 mg of bendrofluazide daily (Fig. 6), and others have shown a similar dose-response relationship with hydrochlorothiazide'261. The increase in plasma urate associated with long-term hydrochlorothiazide is reversed if the dosage is reduced or stopped271. Q- -
-0-5 0
1-25
2-5
50
10-0
1
Bendrofluazide (mg day" ) Figure 3 Effect of 10 weeks administration of graded doses of bendrofluazide on potassium levels (for details see Fig. 1). (Drawn from data, with permission131).
thiazide and related diuretics may enhance calcium reabsorption in thedistal tubuleand reduce urinecalcium19"1 (Fig. 5). Unlike other metabolic actions, this effect is
Hyperglycaemia
Diuretics have been shown to increase fasting glucose levels and impair glucose tolerance curves'28-291. This has been demonstrated in many long-term studies'1"30"32'. The effect on glucose tolerance is usually reversible if the diuretic is stopped'281. Murphy et a/.'331 followed fasting levels and glucose tolerance curves over a period of 14 years and observed a deterioration over this period. In 10 subjects after thiazide withdrawal, tolerance curves improved.
Downloaded from by guest on October 25, 2014
Bendrofluazide (mg day"
usually advantageous. Thiazides have been widely used in idiopathic hypercalciuria but there is now accumulating evidence of a useful anti-osteoporotic effect. Wasnich et a/.'201 reported that patients who have received thiazides had radiologically more dense bones than aged-matched controls, and similarly Ray and colleagues'21>22' found a reduction in the risk of hip fracture in patients over 65 years of age who had used thiazide diuretics. These findings have been confirmed'231 but not by all investigators'241. There seems to be a good case for a prospective clinical trial of thiazides in osteoporosis'221. It should be noted, however, that the administration of slow release frusemide is associated with lower serum calcium than thiazides'251, possibly because calcium re-absorption is inhibited in the ascending loop of Henle.
Adverse reactions to diuretics 99
tn
Adverse reactions to diuretics B. N. C. PRICHARD, C. W. I. OWENS AND A. S. WOOLF
Division of Clinical Pharmacology, University College, London, U.K. KEY WORDS: Thiazides, adverse reactions, hyponatraemia, hypokalaemia, hypercalcaemia, hyperlipidaemia, interactions, potassium retaining diuretics. Diuretics can result in various undesired biochemical changes, such as impotence, skin rashes, nausea, dizziness and lethargy as well as subjective side effects. The side effects are mostly predictable, their effects depending on both the circulatory blood volume and on the transport of water and solute in the renal tubules. Two of the commonest side effects are mild hypovolaemia, when any diuretic is used, and mild hypokalaemia when the non-potassium-sparing diuretics, such as thiazides and frusemide are used. Its occurrence is dose dependent and can be corrected by potassium supplements, but potassium-retaining diuretics, which also correct the often associated fall in serum magnesium, are preferable. Many reports link hypokalaemia with cardiac arrhythmias, but some dispute this association in the absence of the concomitant use ofdigoxin. Hyponatraemia rarely occurs, but can be life threatening. Calcium excretion is markedly reduced, but unlike other electrolyte disturbancesfrom diuretics, this may be valuable: some suggest diuretics have an anti-osteoporotic action. Diuretics increase glucose and insulin resistance and should be used sparingly in diabetics. They rarely cause a non-ketotic hyperosmolar coma. Urate is raised, but clinical gout is not common. Cholesterol elevation has been reported in some studies, but long-term studies indicate that lipid changes are minor. Other rare side effects are not predictable from their pharmacological actions and these include the occurrence of skin rashes, thrombocytopenia, pancreatitis and interstitial nephritis; and ototoxicity from frusemide. Downloaded from by guest on October 25, 2014
Introduction Diuretics have been used in the treatment of hypertension for many years. They possess a wide side effect profile, ranging from subjective effects and idiosyncratic responses to a wide spectrum of metabolic effects. There are also some clinically important interactions with other drugs. Subjective symptoms The MRC1'1 trial in hypertension is the most comprehensive placebo-controlled study, and demonstrated that the most notable side effect from bendrofluazide was the incidence of impotence, which resulted in 19-58 withdrawals from the drug per 1000 patient years. This finding had not hitherto been widely recognised. However, diuretics are not the only antihypertensive drugs associated with impotence; there is a similar incidence with methyldopa and reserpine'21. The patient withdrawals for men with gout were 12-23, for impaired glucose tolerance 9-38 withdrawals per 1000 patient years, for lethargy 6-93, nausea, dizziness or headache 8-56, and constipation 1 -63, all significantly different from placebo. The withdrawals reaching significance for women were, glucose intolerance 601, lethargy 212, nausea, dizziness or headache 16-27, and constipation 212 per 1000 patient years. The MRC1'1 study used what now would be regarded as a high dose of bendofluazide 10 mg. Carlsen et alP] reported a placebo-controlled study of bendrofluazide using l-25mg, 2-5, 50 and lOmg per day with 50-52 patients in each group. The incidence of subjective adverse effects with dosages up to 5 mg were very similar to Correspondence: Prof. B. N. C. Prichard, Division of Clinical Pharmacology, University College London, 5 University Street, London WCIE 6JJ, U.K. 0I95-668X/92/0GOO96 + 08 $08.00/0
1-25
2-5
5-0
10-0
1
Bendrofluazide (mg.day" ) Figure I Effect of 10 weeks administration of graded doses of bendrofluazide on serum creatinine. Placebo n = 52, l-2mg n = 50, 2-5 mg n = 52, 5 0 mg n = 52, 10 mg n = 51. Double-blind parallel group study. (Drawn from data, with permission131). P value, with respect to bendrofluazide 0 mg.
placebo, but moving to 10 mg doubled the total incidence of side effects from 12 out of 52 patients to 24 out of 51 patients. Lowe et al.[A\ in their surveillance study with frusemide, also found a dose-dependent increase in incidence of subjective side effects. Hypovolaemia Hypovolaemia, in its mildest form, is indicated by the small rise in creatinine or urea associated with diuretic administration (Fig. 1)[3], but significant hypovolaemia is much more likely to occur with a loop diuretic. Lowe el a/.'4' found that the incidence of hypovolaemia with frusemide was dose dependent, increasing from 7-5% (n = 347) with doses of 40 mg or less, 17-5% (n = 160) at © 1992 The European Society of Cardiology
Adverse reactions to diuretics 97
Table I
Metabolic side effects of diuretics
1. Hyponatraemia 2. Hypokalaemia 3. Hypomagnesaemia 4. Hypercalcaemia 5. Hyperuricaemia 6. Hyperglycaemia 7. Disturbance lipid metabolism
of commencing thiazide treatment. Besides measures to ensure repletion of sodium in hyponatraemia there is some evidence that magnesium may help to rectify this electrolyte imbalance181. Hypokalaemia and hypomagnesaemia
Downloaded from by guest on October 25, 2014
Hypokalaemia from diuretic administration results from, (i) increased delivery of sodium to the distal tubule so more is available for exchange with potassium (ii) secondary hyperaldosteronism due to volume depletion dosages between 40 and 80 mg, while above 80 mg the and renin secretion and (iii) a shift of potassium into the incidence was 39-7% (n = 78). Hypovolaemia produced cells because of diuretic-induced alkalosis, which results by diuretics may cause a major fall in glomerular fil- from increased bicarbonate resorption by the proximal tration rate (GFR) when used with other drugs such as tubule secondary to hypovolaemia. non-steroidal anti-flammatory agents and angiotensin Carlsen et alP\ in their dose response study of bendroconverting enzyme inhibitors, and this is discussed below. fluazide, found that a near maximum anti-hypertensive effect was observed from l-25mg daily, with similar responses from 2-5, 50 and 10 mg (Fig. 2), whereas there Metabolic side effects was a dose-related fall in potassium levels from 2-5 mg up Diuretics may cause varied metabolic side effects to 10 mg of bendrofluazide (Fig. 3). Similarly, Owens (Table 1). et alP\ in an eight increment dose-response study of xipamide, found the maximum fall in blood pressure was seen with 10 to 20 mg, whereas a fall in potassium only HYPONATRAEMIA Hyponatraemia occurs by a variety of mechanisms as a occurred at 40 mg. The occurrence of hypokalaemia is influenced by result of impairment of free water clearance. The mechanisms include hypovolaemia-induced fall in GFR and potassium intake. Prichard and Brogden'101 reported on enhanced proximal tubule re-absorption: both result in antihypertensive studies of xipamide. If these are the delivery of less fluid to the diluting segment, where arranged according to geography, there is a difference some diuretics, such as thiazides, have the additional between those studies performed in Northern European effect of inhibitory sodium re-absorption. In addition, countries and those where their calorie intake is likely to hypovolaemia will cause both the non-osmotic release of be obtained in a greater proportion from potassium-rich antidiuretic hormone (ADH) and the onset of thirst which fruit and vegetables. While plasma potassium levels can be increased by potassium supplements, Ramsey and his results in water intake. diuretics, Diuretic-induced hyponatraemia is unusual in patients colleagues have shown that potassium-retaining 1 1 11 such as triamterene and spironolactone " , are more under the age of 65 years. Three years into the MRC ' trial as is amiloride'12'131. Additionally, Schnaper there were no important changes in sodium levels in either efficient, 141 men or women, although in both sexes levels compared to et a/.' showed that while potassium-retaining diuretics base line were significantly different from placebo. Small correct potassium, unlike potassium supplements, they increases of 0-9 and 0-6mmol. I"1 were seen in men and also correct magnesium deficiency commonly associated with hypokalaemia (Fig. 4). women respectively. Even common arrhythmias, such as ventricular ectopic There is, however, an important incidence of hyponatraemia in the elderly. Plante and Dessuralt'3' found that beats, are an unfavourable prognostic sign as they are there was an average fall of 141 to 134mmol . 1 " ' from the associated with an increase in the risk of sudden death by administration of 50 mg per day of hydrochlorothiazide a factor of 2-2'131. While some investigations have found for 48 weeks to a group of 19 patients between 65 to 91 years an association between hypokalaemia and cardiac old. Sunderam and Mankikar161 reported 11-3% of hypo- arrhythmias'161, others have not'17181. However, Storstein'151 natraemia incidence (< 130mmol. I"1) in 683 geriatric has shown that some studies indicate an association patients admitted to hospital over a 10 month period, while between hypokalaemia from diuretics and arrhythmias. in 4-5% of the patients the level was under 125mmol .I" 1 . Papdemetriou'181 and his colleagues have found that This was associated with symptoms in these 77 patients of correction of diuretic-induced hypokalaemia from 2-8 to weakness in 28, confusion in 21, postural giddiness in 20, 3-7 mEq . 1 " ' , the administration of diuretics to patients postural hypotension in 17, falls in 13, fits in two, and with left ventricular hypertrophy, even if severe, or transient hemiparesis in one. I n the 49 cases where diuretics diuretics to patients with exercise arrhythmias, had no were responsible for the hyponatraemia, just under a effect on the incidence of arrhythmias. quarter of those were receiving frusemide, as was the case with cyclopenthiazide, while 20% received hydrochlorothiazide amiloride combination. Finally, Ashraf and Hypercalcaemia colleagues'71 described seven patients (Table 2) who Any diuretic may enhance proximal tubular resorption developed dangerous hyponatraemia within a few days of calcium secondary to volume depletion and, in addition,
98 B. N.C.Prichardetal. Table 2
Severe hyponalraemia and thiazides
Age/Sex/In
Na/K
Drug (mg)
Days R.
Time
Outcome
72FBP 50FBP
100/3-1 103/20
10 3
30 h 44h
Paraparesis Coma, died 5/12
75FBP 62MBP
105/1-6 109/2-2
Paralysis 1 Leg Recovered
116/2-9 104/3-0 98/2-4
16 3 7 9 5 5
18h 18 h
68MBP 82FMI 69FBP
HCTZ50 HCTZ50 Metol. 7 HCTZ50 HCTZ50 Metol. 5 HCTZ100 HCTZ25 POLYT.3
14h 48h 56 h
Recovered Coma, died Coma, died 11/52
In = indication; BP = blood pressure; MI = myocardial infarction; Na/K = sodium/potassium levels at presentation in E q . L" 1 ; HCTZ = hydrochlorothiazide; Metol = metolozone; POLYT = polythiazide; Days R = days of treatment before presentation; Time = time in hours required to raise serum sodium to 130 mEq . L~'orabove. After Ashraf etal.m.
Figure 2 Effect of graded doses of bendrofluazide on blood pressure control (for details see Fig. 1). (Drawn from data, with permissionpl)H = week 4; 0 = mean week 10-12.
Hyperuricaemia
An elevation of urate from thiazides is due to reduced blood volume, increased re-absorption by the proximal tubule and to reduced uric acid excretion in the proximal tubule as a result of competition by the diuretic for the organic acid secreting pathway. Carlsen et a/.'31 demonstrated a dose-dependent increase in urate levels from 1 -25 to 10 mg of bendrofluazide daily (Fig. 6), and others have shown a similar dose-response relationship with hydrochlorothiazide'261. The increase in plasma urate associated with long-term hydrochlorothiazide is reversed if the dosage is reduced or stopped271. Q- -
-0-5 0
1-25
2-5
50
10-0
1
Bendrofluazide (mg day" ) Figure 3 Effect of 10 weeks administration of graded doses of bendrofluazide on potassium levels (for details see Fig. 1). (Drawn from data, with permission131).
thiazide and related diuretics may enhance calcium reabsorption in thedistal tubuleand reduce urinecalcium19"1 (Fig. 5). Unlike other metabolic actions, this effect is
Hyperglycaemia
Diuretics have been shown to increase fasting glucose levels and impair glucose tolerance curves'28-291. This has been demonstrated in many long-term studies'1"30"32'. The effect on glucose tolerance is usually reversible if the diuretic is stopped'281. Murphy et a/.'331 followed fasting levels and glucose tolerance curves over a period of 14 years and observed a deterioration over this period. In 10 subjects after thiazide withdrawal, tolerance curves improved.
Downloaded from by guest on October 25, 2014
Bendrofluazide (mg day"
usually advantageous. Thiazides have been widely used in idiopathic hypercalciuria but there is now accumulating evidence of a useful anti-osteoporotic effect. Wasnich et a/.'201 reported that patients who have received thiazides had radiologically more dense bones than aged-matched controls, and similarly Ray and colleagues'21>22' found a reduction in the risk of hip fracture in patients over 65 years of age who had used thiazide diuretics. These findings have been confirmed'231 but not by all investigators'241. There seems to be a good case for a prospective clinical trial of thiazides in osteoporosis'221. It should be noted, however, that the administration of slow release frusemide is associated with lower serum calcium than thiazides'251, possibly because calcium re-absorption is inhibited in the ascending loop of Henle.
Adverse reactions to diuretics 99
tn
