Influence of food and antacid administration on fluoride bioavailability from enteric-coated sodium fluoride tablets

8756-3282189 $3.00 + .OO Copyright 0 1990 Pergamon Press plc

10,401-407 (1989) Printed in the USA. All rights reserved.

Bone,

Influence of Food and Antacid Administration on Fluoride Bioavailability from Enter&coated Sodium Fluoride Tablets P. ARNOLD, M. WERMEILLE,* M. C. CHAPUY,** J. L. SCHELLING and P. J. MEUNIER** Division

of Clinical Pharmacology,

Address

for correspondence

CHUV Lausanne,

Switzerland,

J. BIOLLAZ,

*ZYMA S.A. Nyon,

E. M. GRANDJEAN,*

Switzerland,

and **INSERM

21234 Lyon, France

J&me Biollaz, M.D., Division of Clinical Pharmacology, Vaudois, CH 1011 Lausanne, Switzerland.

and reprints:

Centre Hospitalier Universitaire

Abstract

aluminum hydroxide (Spencer and Lender 1979; Spencer et al. 1980) to eliminate the risk of diminished absorption. These recommendations make the treatment schedule rather complex and might contribute to a decrease in compliance, especially when considering the chronic nature of the treatment and the targeted population, that is, the elderly. Furthermore, a fasting intake of high doses of sodium fluoride has been reported to be associated not infrequently with severe gastrointestinal adverse effects such as nausea, vomiting and peptic ulcer (Rich et al. 1964; Riggs et al. 1980; Riggs et al. 1982). To improve the gastric tolerance of NaF, new galenical preparations have been devised which are resistant to the acid pH of the stomach, allowing the release of fluoride in the small intestine. The bioavailability of these enteric-coated formulations has to be determined and the effect of simultaneous intake of food or drugs studied before new recommendations can be made to simplify the treatment schedule and therefore possibly contribute to improving compliance. The aim of our study was therefore to determine the absorption characteristics and the pharmacokinetics of an enteric-coated sodium fluoride tablet and to assess if and to what extent food and antacids might alter them.

The relative bioavailability of enteric-coated sodium fluoride (NaF) tablets (10 mg F-) has been assessed following admiui&ration with a standard calcium-rich breakfast or calciumpoor lunch, and 2 h before or simultaneously with antacid administration (2.4 g aluminum-magnesium hydroxide), versus intake on an empty stomach. Twelve volunteers were studied 3 thnes according to an open, three-way crossover design over a 24 h period at weekly intervals. Meals were found to decrease the peak serum concentration of NaF from 122 kg/L during f&g (after baseline subtraction) to 71 and 88 kg/L with breakfast and lunch respectively, and to slow its absorption rate with Tmax increasing from 3.3 to 7.3 and 11.2 hours, without altering its bioavailability. Antacid hupaired the bioavailability of NaF by 80% when administered shm&aneously, with AUC decreasing from 987 to 155 pg,h/ L, but had no significant effect when taken 2 h before NaF. In conclusion, the enter&coated NaF tablets used in this study can be administered with food or after a 2-hour delay following antacid administration, but should not be taken simultaneously with antacid. Key Words: Sodium fluoride-Bioavailability-Food-Antacid- Enteric-coated tablets.

Subjects and Methods

Introduction

tielve healthy volunteers, 6 males and 6 females, aged 21 to 43 years (mean 26), weighing between 51 and 76 kg (mean 62) participated in the study. Each volunteer had a medical history taken and underwent a complete physical examination. Routine laboratory tests, including pregnancy testing in female volunteers, were done before and after the administration of the drug. The nature and the purpose of the study had been explained and a written informed consent had been previously obtained. The protocol was approved by the Hospital Ethics Committee.

High doses of sodium fluoride have been reported to increase bone formation (Reutter and Olah 1970; Spencer et al. 1970; Franke et al. 1974; Parsons et al. 1977; Briancon et al. 1981; National Institute of Arthritis 1983; Spencer et al. 1984; Simonen et al. 1985) and to be a useful drug for the treatment of osteoporosis (Spencer et al. 1970; Briancon et al. 1981; Riggs et al. 1982; Spencer et al. 1984) generally in association with a high calcium diet, calcium supplementation and vitamin D (Bernstein et al. 1967; Briancon et al. 1981; Riggs et al. 1982; Spencer et al. 1984). Food and drug interaction have been described with sodium fluoride (NaF) (Ericsson 1958; Cremer et al. 1970; von Patz et al. 1977; Shannon 1977; Jowsey and Riggs 1978; Ekstrand and Ehmebo 1979; Spencer and Lender 1979; Spencer et al. 1980). The recommendation has therefore been made to ingest NaF on an empty stomach and to avoid the simultaneous administration of calcium ions and

Study design

On the study day, the volunteers came to the hospital between 6 and 8 o’clock, after an overnight fast. An intravenous catheter was inserted into an antecubital vein to draw blood. The volunteers received sodium fluoride three times, at weekly intervals, according to a randomized, 401

402

three-way crossover design. The drug was administered at 8 o’clock with 400 mL of tap water. All subjects received the drug once in the fasting state and remained fasting for the following 4 h. Six of them (3 males and 3 females) received the drug once 2 h after the administration of an antacid and once simultaneously with its administration. The remaining six subjects received the drug once with a standard calcium-rich breakfast (2 slices of bread. 10 g of butter, jam, 50 g of cheese and 400 mL of chocolate milk; approximately 900 mg calcium) and once with a standard lunch (steak, french fries, salad, fruit and 400 mL of water: approximately 100 mg calcium) taken at breakfast time. Blood samples for the measurement of fluoride were drawn at 0, IS, 30, 45, 60 and 90 min and at 2. 3, 4. 6, 8. IO. 12 and 24 h from the time of drug administration. Blood was allowed to clot and serum was frozen at - 20°C up to the time of the analysis. Urine was collected by spontaneous voiding at the following time intervals: preceding night-O, O-I, 1-2, 2-4, 4-6,6-12, 12-24 h. To insure adequate and constant urine flow, the volunteers were instructed to drink 150 mL of water per hour, except during the night. Drugs Enteric-coated tablets containing 22.1 mg of sodium fluoride corresponding to 10 mg of the fluoride ion were used for this study (Flurexal@, Zyma, Nyon Switzerland). The dose of antacid we used (40 mL Alucol@ gel. Wander, Switzerland) contained 3 g of aluminum hydroxide, I .4 g of magnesium hydroxide and 3.2 g of sorbitol. Analytical methods Fluoride was measured in serum and urine samples on an ionometer Orion EA 940 with a specific fluoride electrode (Tbsl 1972). The intra and interassay variation coefficients for fluoride measurement in serum were 2%. One mL of the sample was buffered with 0.1 mL of 1-2 cyclohexylenedinitrotetracetic acid for pH adjustment to 5.0-5.5 and aluminum and iron binding. Urine volume was measured by weighing. A hand-held digital pH-meter in automatic temperature compensation mode (model 105 Corning, Halstead, Essex, England) was used for the determination of urinary pH. Calculations and statistical evaluation The kinetic data analysis was performed using GPHARM programme (Gomeni 1984). Estimates of the kinetic parameters were obtained through separate fitting of each subject’s data. For graphic representation however the “naive-averaging-of-data” approach was used. Prior to all calculations, the baseline serum concentration of fluoride was subtracted. The absorption profile of fluoride was determined by the Wagner-Nelson method and the sigmoid model (Gibaldi and Perrier, 1982). The relative bioE, availability (f) was calculated as the ratio of the AU&_, following food or antacid intake and the AUCO_, obtained under fasting conditions. The elimination rate constant (B) was calculated from the least-squares tit of the terminal portion of the log plasma concentration-time curve and the elimination ty, was calculated as t, = 0.693/B. Renal clearance (CL,) was calculated using both serum drug concentration (C,) and cumulative amount excreted in the urine (A,) according to the following model:

P. Arnold et al.: Food and antacid intake: effect on F- availability C,

=

2 Ai

exp(-“4

A, = 2 CL, A, (1 - exp(mal”) 1 N,

Renal clearance was calculated by linear regression. A t test was performed to compare the computed intercept to the expected theoretical value (zero). If the linear regression was statistically significant (p < 0.05) and the intercept not significantly different from zero, the simultaneous fitting of plasma and urine data was performed using the above model and the extended least squares procedure. The significance of differences between means was evaluated by analysis of variance (ANOVA) and, when appropriate, treatment levels were compared by means of Fisher’s Least Significant Difference Test (Snedecor and Cochran, 1974). Results are given as mean ‘- SEM.

Results Table I summarizes the characteristics of the subjects studied. The mean weight of the male subjects attributed to the “meal” group was higher than that of the “antacid” group (75.1 versus 62.3 kg). The weight of the female subjects was similar in both groups (56.3 versus 56.7 kg). The mean baseline serum concentration of fluoride was 33.8 I+_ 1.7 and 35.1 ? 2.0 pg/L in the “meal” and “antacid” groups respectively. Ejfect of food The individual absorption

profiles of fluoride are shown in started after a mean time lag of 1.3 * 0.1 h and the mean time for 50% absorption was 2.5 * 0.4 h. When taken with breakfast (middle panel) and lunch (lower panel), the absorption process was delayed and more variable, with a time lag of 3.2 ? 0.7 and 3.4 rf~1.1 h and a mean time for 50% absorption of 6.2 t 0.9 and 6.9 * 2.2 h respectively. Fig. 2 (upper panel) shows the mean serum concentration time curve obtained in subjects under fasting conditions and when fluoride was administered concomitantly with food. When fasting, the mean of the concentrations at the peak reached 122 + 35 kg/L above baseline and was achieved at a mean time of 3.3 ? 0.6 h (Table II). When Fig. 1. When fasting (upper panel), the absorption

Table I. Details of subjects in the experiment. Subject

Age

Body weight

Height

Sex

[years]

[Kg1

[cm1

M M M F F F

26 30 27 23 21 44

76 14 75 54 57 58

183 183 180 163 174 166

M M M F F F

28 26 25 22 27 23

57 60 70 51 58 61

174 182 189 167 168 155

“Meal” group RI BR BH BC SI WM “Antacid” SC

CJR ZN SJ LP SJ

group

403

P. Arnold et al.: Food and antacid intake: effect on F- availability

.-.

Fasting .-.Brwakfart l-*Lunch

0

4

8

12 TIME

16

20

24

hours

5000 9 B k 4000 E 0 0

4

6

12

16

20

24

5 3000 s Y i 2000 3 2 &!lOOO d 2 0 0

-•

l Famtlng A--.Bnakfad

l--=Lunch

4

8

12 TIME

0

4

6

12 TINE

16

20

24

houn

Fig. 1. Percent absorbed-time plot after a single enteric-coated 10 mg F- tablet. The absorption profiles are the result of administering the drug to 6 subjects either fasting (top panel), with a breakfast (middle panel) or with a lunch (bottom panel). Each line in figures represents a subject.

&ax dmL

6 6 6

122 f 35 71 * 11 88 * 25

6 6 6

116 f 21 59 f 15t 138 f 26

Tmax h

20

24

Fig. 2. Top panel: Average fluoride concentrations in serum of 6 healthy subjects who received single oral doses of an entericcoated tablet of 10 mg F- under fasting conditions (circles), simultaneously with breakfast (triangles) or lunch (squares). The baseline serum levels of F- have been subtracted from the measured values. Bottom panel: Average cumulative amounts of F- excreted in the urine of the same subjects.

Table II. Kinetic parameters obtained following the ingestion of an enteric-coated n

16 hours

TV2 h

NaF tablet (10 mg F-).

A&,_, pg . h/L

f

ACu

mg/24 h

“Meal” group Fasting Breakfast Lunch “Antacid”

3.1 2 0.4 5.9 2 1.9 5.8 +- 1.3

766 + 162 717 2 8 737 ‘- 0

0.94 0.96

-

4.1 + 0.4 2.9 * 0.3 3.6 2 0.3

5.4 * 1.3 1.3 * 03** ’ t 4.7 f 0:9

987 f 121 155 * o***,ttt 838 2 142

O.l6*=,ttt 0.89

4.4 t 0.5 1 8 ? 0 3** ’ $ 4:1 * 0:7

group

Fasting Antacid T,, Antacid T - 2 Mean *p < t p < c _:

3.3 2 0.6 7.3 2 0.8 11.2 * 4.1

4.3 * 0.6 0.9 2 0.1***,ttt 4.0 2 1.1

f SEM. 0.05; **p < 0.01; ***p < 0.001, versus fasting. 0.05; $p < 0.01; http < 0.001, versus antacid T - 2. peak concentration. T_: time to peak concentration. Tm: plasma half-life. AUC,_,: area under the serum concentration time curve from time 0 to infinity. f:relative bioavailability. amount excreted in the urine during the 24-hour period. A,:

404

sodium fluoride was given with breakfast and lunch, the mean peak concentrations were also reduced (71 * 12 l.~g/L and 88 * 25 kg/L respectively) and the mean time to reach them increased (7.3 t 0.8 h and 11.2 ? 4.1 h respectively) (Table II). The AUCs following food administration were not significantly different from the AUCs obtained under fasting conditions. The relative bioavailability of sodium fluoride tablets taken with food was slightly less than unity (Table II). However, all the above changes induced by food intake were not statistically significant. On the lower panel of Fig. 2, the cumulative urinary excretion of fluoride, obtained under fasting conditions and with food administration, is plotted against time. Due to the delayed absorption, a small, not statistically significant difference was observed in the amount excreted during the 24-hour period when sodium fluoride was taken with food as compared with intake in a fasting state. The amount excreted during the 24-hour periods represented 41% of the amount ingested when fasting, 36% with lunch and 29% with breakfast. The relative bioavailability of sodium fluoride calculated from urinary data was not significantly altered by food. The mean renal clearance of fluoride was 73,50 and 60 mL/min following drug intake on an empty stomach or with breakfast and lunch respectively.

P. Arnold et al.: Food and antacid intake: effect on F- availability

0

oh. 0

4

6

.‘...‘...“‘.“.‘,...I

12

16

20

24

4

6

12

16

20

24

4

6

12

16

20

24

Effect of antacid

In Fig. 3 are shown the individual absorption profiles of fluoride following drug intake without antacid (upper panel), simultaneously with it and 2 h after intake. Simultaneous antacid intake strikingly changed the absorption process of fluoride. No time lag could be determined in most subjects. The time to 50% absorption was decreased compared to fluoride intake without or with antacid intake 2 h before (0.4 -t 0.1 versus 3.1 * I .O (p < 0.05) and 2.3 + I .O h respectively). The upper panel of Fig. 4 indicates the time course of the mean serum concentrations of fluoride when the sodium fluoride tablet was taken in a fasting state, 2 h following intake of the antacid or simultaneously with it. In the last mentioned situation, the mean of the serum concentrations at the peak decreased from 116 ? 21 (no antacid) to 59 2 15 t&L, and the mean time to reach them from 4.5 +- 1.2 to 0.9 + 0.1 h, and the mean serum half-life of fluoride was significantly shortened (Table II). AUC was markedly altered, falling from 987 + 141 during fasting to 155 ? 50 pg.h/L with simultaneous antacid intake. Consequently the relative bioavailability fell, with only 16% of fluoride being absorbed as compared with fasting. When sodium fluoride was taken 2 h after the antacid, the serum concentration profile and the bioavailability did not differ significantly from the fasting situation (Table II). The cumulative urinary excretion of fluoride was markedly decreased by the simultaneous antacid intake, as shown on the lower panel of Fig. 4. The intake of antacid 2 h before did not influence the total amount excreted although there was a trend towards a faster excretion rate, reflecting the slightly faster rate of absorption. The total amount excreted, while fasting, during the 24 h period represented 44% of the amount ingested. This percentage was 41% when NaF was ingested 2 h following the antacid administration and only 18% when it was taken simultaneously with the antacid (Table II). The renal clearance of fluoride was 58, 58 and 82 mL/min following intake on an empty stomach, simultaneously with or 2 h after antacid intake respectively.

0

NYE

1

houn

Fig. 3. Percent absorbed-time plot after a single enteric-coated 10 mg F- tablet. The absorption profiles are the result of administering the drug to 6 subjects either fasting (top panel), with an antacid (middle panel) or 2 h after antacid administration (bottom panel). Each line in figures represents a subject.

The hourly urinary volume did not differ significantly between all the study phases (Figs. 5 & 6, upper panels). The urinary pH was significantly higher during part of the day when food was given, as compared to the fasting state, while antacid administration did not alter it (Figs. 5 & 6, lower panels).

Fluoride absorption is a passive process which takes place in the stomach and the small intestine (Cremer and Biittner, 1970). Passive absorption is facilitated when a drug is nonionized. The absorption of fluoride, an acid with a pKa of 3.4, is therefore optimum in the acidic milieu of the stomach (Whitford and Pashley, 1984). Consequently, fluoride is rapidly and nearly completely absorbed from an empty stomach (Cremer and Biittner 1970; Ekstrand et al. 1978). However fasting intake of high doses of fluoride is associated with side effects such as gastric pain and nausea (Kanis and Meunier 1984). These adverse effects are reduced by concomitant administration of the drug with food or antacids (Rich et al. 1964), which may however reduce its bioavailability (Spencer et al. 1980). The gastrointestinal side effects of fluoride might be de-

F! Arnold et al.: Food and antacid intake: effect on F- availability -I

r

280 z Y 200 3 d > 150 g 3

1; % I i:

l~.“,“‘,‘.“.‘.,“‘~“” 0

4

8

100

50 0

12 16 TIME hourt

20

24

COUECTIDNPERIOD

(h)

6 i

s l

-• Fotilng

.-a 0

AntacId

TO

J 4

8

12

16

20

24

TIME hours

0

-0

O-l

l-2

COU.ECTlON

2-4

4-6

PERIOD

6-12

12-24

(h)

Fig. 4. Top panel: Average fluoride concentrations in serum of 6 healthy subjects who received single oral doses of an entericcoated tablet of 10 mg F- under fasting conditions (circles), simultaneously with antacid intake (triangles) or 2 h after antacid intake (squares). Bottom panel: Average cumulative amounts of F- excreted in the urine of the same subjects.

Fig. 5. Hourly urinary volume (top panel) and urinary pH (bottom panel) measured following the administration of a IO mg F- enteric-coated tablet under fasting conditions (blank box), with breakfast (left diagonal) or with lunch (right diagnoal). (* and + denote a statistical difference at p < 0.05 versus time 0 and versus the corresponding fasting period, respectively.)

creased by coating sodium fluoride with an acid resistant layer (Franke 1978). This enteric coating is expected to in-

centration, without altering its elimination kinetics and bioavailability, as measured by the AUCs. These findings are best explained by the delay in gastric emptying caused by food intake. The large variability in the time to peak we observed with meals may also be related to the large variability of the gastric emptying rate which follows food administration (Nimo, 1976). The calcium content of the food did not influence the bioavailability, and this is consonant with the results obtained following the administration of fluoride simultaneously with calcium gluconolactate and carbonate (Briancon et al. 1988). Ingestion of an enteric-coated NaF tablet simultaneously with antacid markedly changed its absorption and elimination kinetics. The disappearance of any time lag to the onset of absorption is probably a consequence of the increased gastric pH which altered the enteric coating. Furthermore, the higher gastric pH should increase the amount of ionized fluoride; this could increase the formation of insoluble salts of magnesium and aluminum (Hodge 1961) and decrease the bioavailability of fluoride. The interaction between antacid and sodium fluoride is, however, limited in time, since a two-hour delay between their respective intake is not associated with any relevant interaction. Although our study was done in young subjects, it has been shown that, in the absence of concomitant diseases or drug administration, the rate but not the extent of the ab-

troduce a time lag in the absorption process, as well as to lower the absorption rate and consequently the peak serum concentration, since in the alkaline pH of the small intestine, fluoride will have to be absorbed essentially under the ionized form. The time lag between ingestion and the onset of the absorption was approximately 1 h and the peak serum concentration was lower than what has been reported with a similar dose of fluoride given as a “plain” preparation. However, the AUCs were similar to that obtained with “plain” preparations (Ekstrand et al. 1977) suggesting bioequivalence of this enteric-coated formulation. Qualitatively similar results have been obtained with other enteric-coated preparations (Lorent et al. 1979; Chaleil et al. 1988). The observed decreased bioavailability following enteric-coated tablets administration reported in one study (Hasvold and Ekren 1981) is probably due to a too short study time and reflects more a delayed than a decreased absorption. However, biological equivalence does not mean therapeutic equivalence. It remains to be determined if lower peak serum concentrations with similar AUCs present advantages or disadvantages with regard to therapeutic or adverse effects. When enteric-coated NaF tablets were ingested with meals, a greater time lag and a slower absorption rate were found, leading to a further decrease in the peak serum con-

406

P. Arnold et al.: Food and antacid intake: effect on F- availability should be recommended if both drugs have to be administered to the same patient.

References

O-l

l-2

2-4

COLLECTION

4-I

6-12

PERIOD

(h)

12-24

Bernstein, D. S.; Cohen. P Use of sodium fluoride in the treatment of osteoporosis. J. C/in. Endocrinol. Merab. 27:197-210; 1%7. Briancon. D.; Meunier. P. J. Treatment of osteoporosis with fluoride, calcium, and vitamin D. Orthop. Clin. North Am. 12(3):629-648; 1981. Briancon, D.: Quillet. P.; Duplan, B.; Chapuy, M. C.; Arlot, M.; Meunier. P. .I. Comparaison de la biodisponibilite du fluor a la suite de I’adminis[ration de fluorure de sodium seul ou en association avec du calcium. ThPrapie 43:f07- 110: 1988. Chaleil. D.; Boulard, C.: Guillaume, M.; Allain. P. Etude pharmacociuetiqne du fluorure apres une administration unique al’Osteofluor@. Therupie 43%7;

COLLECTION

PERIOD

(h)

Fig. 6. Hourly urinary volume (top panel) and urinary pH (bottom panel) measured following the administration of IO mg F- entericcoated tablet under fasting conditions (blank box), simultaneously with (left diagonal) or 2 h following antacid intake (right diagonal). (*denotes a statistical difference at p < 0.05 versus time 0.)

sorption

changes with age (Nimo 1976; Stevenson et al. 1979). The gastric pH may be increased in some elderly

patients, and this could theoretically mimic the effect of antacid on the enteric coating. However, in the absence of magnesium and aluminum salts, only the absorption rate would be altered. It has been shown that, at steady state, urinary output of fluoride is about 50% of the dose given. In our study, the urinary recovery of fluoride was approximately 40% of the dose administered. These results are in agreement with previous studies using enteric-coated formulations (Hasvold and Ekren 1981; Chaleil et al. 1988). This suggests that the availability of the enteric-coated formulation used in our study was equivalent to that of “plain” sodium fluoride tablets. This should be confirmed by further studies in which the kinetics of fluoride is determined under steady state conditions with “plain” and “entericcoated” preparations. In conclusion, food changes the absorption without altering the elimination profile or the bioavailability of the enteric-coated preparation used in this study. The treatment schedule may therefore be simplified, that is, the drug can be taken with food. This could improve the compliance of the patients. However, such enteric-coated tablets should not be given simultaneously with antacids since their bioavailability is markedly reduced; a time interval of at least 2 h between fluoride and antacid intake

1988.

Cremer. H. D.; Buttner, W. Absorption of fluorides. In: Fluorides and human hralth. Geneva: WHO.; 1970:~. 75-91 (Chapter 3). Ekstrand. J.; Alvan. G.; Boreus, L. 0.; Norlin. A. Pharmacokinetics of fluoride in man after single and multiple oral doses. Eur. J. Chin. Pharmacol. 12:311-317; 1977. Ekstrand, J.; Ehmebo, M. InfIuence of milk products on fluoride bioavailability in man. Eur. J. C/in. Pharmacol. 16:211-215; 1979. Ekstrand. J.: Ehmebo, M.; Boreus, L. 0. Fluoride bioavailability after intravenous and oral administration: importance of renal clearance and urine flow. Cfin. Pharm. Thu. 23:329-337; 1978. Ericsson, Y. The state of fluoride in milk and its absorption and retention when administered in milk. Investigations with radioactive fluorine, Acta Odor&. Stand. 16:51-77, 1958. Franke. J. Our experience in the treatment of osteoporosis with relatively low sodium fluoride doses. In: Courvoisier. B.; Donath, A., eds. Fluoride and Bone. Bern: H. Huber; 1978:~. 256-262. Franke, J.: Rempel, H.; Franke, M. Three years’ experience with sodiumfluoride therapy of osteoporosis. Acta Orthop. Sand. 45: I-20; 1974. Gibaldi, M.: Perrier, D. Pharmacokinetics. New York: Marcel Dekker Inc.,; 1982. Gomeni, R. PHARM: an interactive graphic program for individual and population pharmacokinetic parameters estimation. Camp. Biol. Med. 14:25-34; 1984. Hasvold. 0.; Ekren, T. In vitro release and in viva serum fluoride levels and urinary excretion after different sodium fluoride tablets. Eur. J. C/in. Pharmacol. 19:225-230; 1981. Hedge. H. C. Metabolism of fluorides. JAMA 177:313-316; 1961. Jowsey, J.; Riggs, B. L. Effect of concurrent calcium ingestion on intestinal absorption of fluoride. Metabolism 27~971-974; 1978. Kanis, J. A.; Meunier, I? J. Should we use fluoride to treat osteoporosis? A review. Quant. J. Med. 210:145-164; 1984. Lorent, M.; Gervois, J. P.; Sondagh, J.; Dodion, L. Etude de la biodisponibilite d’une preparation de comprimses enterosolubles de fluorure de sodium. J. Pharm. Belg. 34~272-278; 1979. National Institute of Arthritis, Diabetes, and Digestive and Kidney Disease. Osteoporosis: cause, treatment, prevention. National Institutes of Health Publication no. 83-2226. Bethesda, MD: NIH; 1983:~. l-7. Nimo, W. S. Drugs, disease and gastric emptying. Clin. Phormacokinet. l:lO!z-203;

1976.

Parsons, V.; Mitchell, C. J.; Reeve, J.; Hesp, R. The use of sodium fluoride, vitamin D and calcium supplements in the treatment of patients with axial osteoporosis. C&if. Tissue Int. 22:236-240; 1977. Reutter, F. W.; Olah, A. J. Bone biopsy findings and clinical observations in long-term treatment of osteoporosis with sodium fluoride and vitamin Dr. In: Vischer, T. L.. ed. Huoride in medicine. Bern: Hans Huber; 1970:~. 249-255. Rich, C.; Ensink, J.; Ivanovich, P. The effects of sodium fluoride on calcium metabolism of subjects with metabolic bone diseases. J. Clin. Invest. 43~545-556; 1964. Riggs, B. L.; Hodgson, S. F.; Hoffman, D. L.; Kelly, P. J.; Johnson, K. A.: Taves, D. ‘Beatment of primary osteoporosis with fluoride and calcium. Clinical tolerance and fracture occurrence. JAMA 243:446-449; 1980. Riggs, B.; Seeman, E.; Hodgson, S. F.; ‘Ihves, D. R.; O’Fallon, W. M. Effect of the fluoride/calcium regimen on vertebral fracture occurrence in postmenopausal osteoporsis. Comparison with conventional therapy. N. Engl. J. Med. 306:446-450; 1982.

P. Arnold

et al.: Food and antacid

intake:

effect on F- availability

Shannon, I. L. Biochemistry of fluoride in saliva. Caries Res. ll(Suppl. 1):206-225; 1977. Simonen. 0.; Laitinen, 0. Does fluoridation of drinking-water prevent bone fragility and osteoporosis? Lancelii:432-434; 1985. Snedecor, G. W.; Cochran, W. G. Statistical methods (6th ed.). Ames: The Iowa State University Press; 1974. Spencer, H.; Kramer, L.; Norris, C.; Wiatrowski, E. Effect of aluminum hydroxide on fluoride metabolism. C/in. Pharm. Thu. 28:529-535; 1980. Spencer, H.; Kramer, L.; Wiatrowski, E.; Lender, M. Fluoride therapy in metabolic bone disease. Isr. J. Med. Sci. 20:373-380; 1984. Spencer, H.; Lender, M. Adverse effects of aluminum-containing antacids on mineral metabolism. Gasrroenrerology 76:606-616; 1979. Spencer, H.; Lewin, I.; Osis, D.; Samachson, .I. Studies of fluoride and calcium metabolism in patients with osteoporosis. Am. J. Med. 49:814822; 1970.

407 Stevenson, I. H.; Salem, S. A. M.; Shepherd, A. M. M. Studies on drug absorption and metabolism in elderly. In: Crooks, J.; Stevenson, I. H., eds. Drugs and the elderly-perspectives in geriatric clinical pharmacology. London: The MacMillan Press Ltd.; 1979:~. 51-64. ‘Ihsl, _I.Fluoride ion activating electrode as a suitable mean for exact determination in urinary fluoride. Anal. Chem. 44:1693-1694; 1972. von Patz, J.; Henschler, D., Fickenscher, H. Biovertiigbarkeit von fluorid aus verschiedenen salzen und unter dem Einfluss verschiedener Nahrungsbestandteile. Dtsch. zahniirzl. Z. 32:482-486; 1977. Whitford, G. M.; Pashley, D. H. Fluoride absorption: the influence of gastric acidity. C&if. Tissue Int. 36:302-307; 1984

Received: September 15, 1987 Revised: October 8, 1988 Accepted: August 14, 1989