Singh N, Singh PN, Hershman JM. Effect of Calcium Carbonate on the Absorption of Levothyroxine. JAMA. 2000;283(21):2822-2825. doi:10.1001/jama.283.21.2822
Author Affiliations: Division of Endocrinology and Metabolism, VA Greater Los Angeles Healthcare System, UCLA School of Medicine, Los Angeles, Calif (Drs N. Singh and Hershman), and Division of Epidemiology, Loma Linda Medical Center, Loma Linda, Calif (Dr P. Singh).
Context The effect of calcium carbonate on the absorption of levothyroxine has
not been studied systematically. Such a potential drug interaction merits
investigation because concurrent treatment with both drugs is common, particularly
in postmenopausal women.
Objective To investigate the potential interference of calcium carbonate in the
absorption of levothyroxine.
Design Prospective cohort study conducted from November 1998 to June 1999,
supplemented with an in vitro study of thyroxine (T4) binding to
Setting Veterans Affairs Medical Center in West Los Angeles, Calif.
Patients Twenty patients (age range, 27-78 years; n=11 men) with hypothyroidism
who were taking a stable long-term regimen of levothyroxine were included
in the study. All patients had serum free T4 and thyrotropin values
in the normal range before beginning the study.
Intervention Subjects were instructed to take 1200 mg/d of elemental calcium as calcium
carbonate, ingested with their levothyroxine, for 3 months.
Main Outcome Measures Levels of free T4, total T4, total triiodothyronine
(T3), and thyrotropin, measured in all subjects at baseline (while
taking levothyroxine alone), at 2 and 3 months (while taking calcium carbonate
and levothyroxine), and 2 months after calcium carbonate discontinuation (while
continuing to take levothyroxine).
Results Mean free T4 and total T4 levels were significantly
reduced during the calcium period and increased after calcium discontinuation.
Mean free T4 levels were 17 pmol/L (1.3 ng/dL) at baseline, 15
pmol/L (1.2 ng/dL) during the calcium period, and 18 pmol/L (1.4 ng/dL) after
calcium discontinuation (overall P<.001); mean
total T4 levels were 118 nmol/L (9.2 µg/dL) at baseline,
111 nmol/L (8.6 µg/dL) during the calcium period, and 120 nmol/L (9.3
µg/dL) after calcium discontinuation (overall P=.03).
Mean thyrotropin levels increased significantly, from 1.6 mIU/L at baseline
to 2.7 mIU/L during the calcium period, and decreased to 1.4 mIU/L after calcium
discontinuation (P=.008). Twenty percent of patients
had serum thyrotropin levels higher than the normal range during the calcium
period; the highest observed level was 7.8 mIU/L. Mean T3 levels
did not change during the calcium period. The in vitro study of T4
binding to calcium showed that adsorption of T4 to calcium carbonate
occurs at acidic pH levels.
Conclusions This study of 20 patients receiving long-term levothyroxine replacement
therapy indicates that calcium carbonate reduces T4 absorption
and increases serum thyrotropin levels. Levothyroxine adsorbs to calcium carbonate
in an acidic environment, which may reduce its bioavailability.
Levothyroxine sodium is commonly prescribed for the treatment of hypothyroidism
and thyroid neoplasia. The absorption of levothyroxine is approximately 80%
after oral administration.1,2
Certain drugs have been shown to interfere with the absorption of levothyroxine.
These include ferrous sulfate,3 sucralfate,4,5 bile acid sequestrants used to treat
hypercholesterolemia,6 and aluminum hydroxide
given as an antacid.7,8 In addition,
high-fiber diets may impair thyroxine (T4) absorption,9 and in some cases, food may delay or impair levothyroxine
absorption.10 Other drugs may accelerate the
disposal of T4 and thus increase the dose requirement; these include
phenytoin (Dilantin),11 carbamazepine (Tegretol),11 and sertraline (Zoloft).12
Calcium carbonate is taken by postmenopausal women for prevention or
therapy of osteoporosis. In general, the use of calcium carbonate is increasing
because of concern about osteoporosis. The largest group of patients taking
T4 is postmenopausal women. Calcium carbonate has been shown to
prevent osteoporosis induced by thyrotropin-suppressive doses of levothyroxine
in postmenopausal women.13
There is concern that calcium carbonate may reduce the absorption of
levothyroxine. Although there are anecdotal claims to this effect, a MEDLINE
search revealed no published prospective research studies of this potentially
important interaction. Therefore, we studied the potential interference of
calcium carbonate in the absorption of levothyroxine. A study was performed
in 20 patients with hypothyroidism. In addition, a study of T4
binding to calcium carbonate was performed in vitro.
Study Population. Study subjects were recruited at the Veterans Affairs Medical Center
in West Los Angeles and at UCLA Medical Center from patients in the endocrinology
clinics. Twenty patients (11 men, 9 women), ages 27 to 78 years, with hypothyroidism
secondary to Hashimoto thyroiditis, surgical thyroidectomy, or radioiodine
ablation of the thyroid who took levothyroxine in a dose of 1.0 µg/kg
or greater were selected for study. They had an initial screening visit, which
consisted of medical history and physical examination. All patients had normal
free T4 values and serum thyrotropin levels in the normal range
of 0.6 to 4 mIU/L. Patients taking the following medications were excluded:
aluminum hydroxide antacids, iron preparations, sertraline or similar drugs,
phenytoin, carbamazepine, colestipol or cholestyramine, and fiber-supplemented
diets. Study and consent procedures were reviewed and approved by the institutional
review board of the Veterans Affairs Greater Los Angeles Healthcare System,
and each patient gave informed, written consent.
Data Collection. Patients had initial baseline measurement of thyroid function tests.
If the initial free T4 and thyrotropin levels fell within the normal
range, they were maintained on the prestudy dose of levothyroxine. If the
initial free T4 and thyrotropin levels were not normal, then the
dose of levothyroxine was adjusted. Patients were advanced to the next stage
of the study once they had normal free T4 and thyrotropin levels
after at least 6 weeks on a stable dose of levothyroxine. The dose of levothyroxine
was not changed during the remainder of the study. The levothyroxine preparation
used throughout the study was Synthroid.
Each patient then was given 1200 mg of calcium (as calcium carbonate,
Goldmine brand) with instructions to ingest it daily with the levothyroxine
in the morning on an empty stomach. The patients returned for follow-up measurement
of serum free T4, total T4, total triiodothyronine (T3), and thyrotropin at 2 months and 3 months after beginning calcium.
The calcium carbonate was discontinued at the 3-month follow-up visit and
levothyroxine was continued. The patients returned 2 months after discontinuing
calcium for thyroid function tests. Pill counts, brief histories, and physical
examinations were carried out at each visit. The patients did not take the
calcium and levothyroxine before the blood was sampled on the day of the visit.
Most blood samples were collected during the morning.
Laboratory Studies. Free T4 and total T4 levels were determined by
the Coat-A-Count radioimmunoassay methods (Diagnostic Products Corporation,
Los Angeles, Calif). The T3 level was determined by radioimmunoassay
(Magic T3 Radioimmunoassay, Bayer Corporation Diagnostic Division,
Tarrytown, NY). Measurement of serum thyrotropin was done by immunoassay with
3 monoclonal antibodies (Nichols Institute Diagnostics, San Juan Capistrano,
Statistical Analyses. Mean values for the thyroid function tests (free T4, total
T4, total T3, and thyrotropin) for the initial visit
while taking levothyroxine, 2 visits while taking calcium plus levothyroxine,
and the final visit while taking levothyroxine were compared using the repeated
measures multivariate analysis of variance test. P
values from this test were computed for the overall difference between the
groups (baseline levothyroxine vs 2 visits while taking levothyroxine plus
calcium vs final levothyroxine) and for specific contrasts between the groups
(baseline levothyroxine vs 2 visits while taking levothyroxine plus calcium,
final levothyroxine vs 2 visits while taking levothyroxine plus calcium).
Timeline. The timeline for the study was as follows: (1) 0 months: baseline taking
levothyroxine, add calcium at visit; (2) 2 months: visit No. 1 taking levothyroxine
plus calcium; (3) 3 months: visit No. 2 taking levothyroxine plus calcium,
discontinue calcium at visit; and (4) 5 months: final visit taking levothyroxine,
after calcium discontinuation.
An in vitro study of T4 binding to calcium was modified after
that of Liel et al.9 The buffer solution consisted
of 0.1% bovine serum albumin, phosphate-buffered saline, and T4
(at a concentration of 0.8 µg/dL). Four hundred milligrams of calcium
carbonate (Sigma, reagent grade) was added to 1 mL of the buffer solution
and serially diluted (with buffer solution containing 0.1% bovine serum albumin,
phosphate-buffered saline, and cold T4) to achieve the following
concentrations: 400 mg/mL, 100 mg/mL, 25 mg/mL, 6.25 mg/mL, 1.56 mg/mL, and
0.39 mg/mL. Ten microliters of 125I-T4 (NEN Life Science
Products, Boston, Mass, specific radioactivity 5700 µCi/µg, 10
µL containing 25,000 to 30,000 cpm) was added to the serial dilutions
of calcium carbonate as well as to a buffer solution without calcium. Tubes
were incubated in a shaking bath for 2 hours at 37° C. At the end of the
incubation period, samples were centrifuged at 1000 g
for 10 minutes. Two hundred microliters of supernatant was transferred to
a second set of tubes and tubes were counted for 2 minutes in a gamma well
Adsorption to calcium carbonate was examined by calculating the percentage
change in 125I-T4 in the supernatant (in cpm) in serial
dilutions of calcium carbonate compared with buffer plus 125I-T4 alone. Four trials were carried out with the buffer adjusted to a
pH of 7.4. Five trials were carried out with the buffer adjusted with hydrochloric
acid to a pH of 2.0 to simulate gastric acidity.
The mean free T4 level was significantly reduced as a result
of calcium treatment from a baseline of 17 pmol/L (1.3 ng/dL) to 15 pmol/L
(1.2 ng/dL) during the calcium period, and increased to 18 pmol/L (1.4 ng/dL)
after calcium discontinuation (overall P<.001).
The mean total T4 level was also significantly reduced as a result
of calcium treatment from 118 nmol/L (9.2 µg/dL) at baseline to 111
nmol/L (8.6 µg/dL) during the calcium period, and increased to 120 nmol/L
(9.3 µg/dL) after calcium discontinuation (overall P=.03) (Table 1). The mean
serum thyrotropin level increased significantly from 1.6 mIU/L to 2.7 mIU/L
with calcium treatment, and then dropped to 1.4 mIU/L after calcium discontinuation
(overall P=.008) (Table 1). Mean T3 did not change as a result of calcium
Thirteen of 20 patients had a reduction in free T4 during
the calcium phase, and 7 patients had no change (Figure 1, left). Thirteen of 20 patients had an increase in thyrotropin
level during the calcium phase and in 4 patients, it rose above the normal
range. Four of the remaining 7 patients had no substantial change in thyrotropin
level and 3 had a slight decrease (Figure
1, right). Pill counts were carried out at each visit and confirmed
In the in vitro study in which 125I-T4 was incubated
with serial dilutions of calcium carbonate, at pH 7.4 the fraction of 125I-T4 recovered in the supernatant after calcium was added
was not different from 125I-T4alone. The mean percentage
of added T4 recovered in the supernatant at pH 7.4 for different
concentrations of calcium carbonate ranged from 97% to 109% (Figure 2). At pH 2.0, however, over 5 trials, the mean percentage
of 125I-T4in the supernatant was reduced to 52% at 400
mg/mL of calcium carbonate and to 90% at 100 mg/mL (Figure 2). At 400 mg/mL and 100 mg/mL, calcium carbonate was present
in a slurry rather than a clear solution.
The results of the study of the 20 patients with hypothyroidism receiving
T4 replacement therapy indicate that calcium carbonate has a modest,
but significant, effect on thyroid function, most likely due to blocking the
absorption of levothyroxine. The administration of calcium and levothyroxine
in these patients was associated with a significant reduction in mean serum
free T4 and total T4 levels during the calcium period.
The increase of these values in most patients after calcium discontinuation
strengthened the likelihood that the changes were due to calcium ingestion.
The effect of calcium on thyrotropin level was more dramatic than that
on free T4 and total T4. The mean thyrotropin level
increased significantly from 1.6 to 2.7 mIU/L with calcium treatment, and
then dropped to 1.4 mIU/L after calcium discontinuation (overall P=.008). Thirteen of 20 patients had an increase in thyrotropin level
during the calcium phase. The serum thyrotropin level was above the normal
range in 4 (20%) of 20 patients, with the highest observed level being 7.8
mIU/L. These 4 patients would have required an increased dose of levothyroxine
if they continued to take it with calcium carbonate. Mild thyrotropin elevation
indicates subclinical hypothyroidism, in this case due to inadequate replacement
therapy. Thyroid hormone treatment of patients with subclinical hypothyroidism
may improve lipid profiles and symptoms. 14,15
Since the study was not placebo controlled, it is possible that some
differences in the thyroid function tests were due to changes in compliance
in taking the medication. The study design, however, used each patient as
his or her own control and showed that the effects were reversible when calcium
was discontinued. In addition, pill counts provided evidence for patient compliance.
The results of the study were similar to the clinical observations made
by Schneyer in 1998.16 However, whereas this
study demonstrated a modest influence of calcium ingestion on serum thyrotropin
values, the Schneyer data suggest a more profound effect. Schneyer reported
that in 3 women with thyroid cancer receiving levothyroxine suppression therapy,
the simultaneous ingestion of calcium carbonate and levothyroxine decreased
the efficacy of T4. The first patient took 1200 mg of calcium (in
the form of Tums) and her thyrotropin level rose from 0.08 mIU/L at baseline
to 13.2 mIU/L, dropping to 0.6 mIU/L after calcium discontinuation. The second
and third patients took 1000 mg of calcium (in the form of Os-Cal) and had
similar trends in thryotropin level. The Schneyer data suggest that the effect
of calcium on levothyroxine efficacy could be avoided by dosing calcium separately
(approximately 4 hours) from T4.
Liel et al7,8 demonstrated
the nonspecific adsorptive capacity of aluminum hydroxide for T4.
The in vitro experiment paralleled a significant increase in serum thyrotropin
level in patients given aluminum hydroxide and levothyroxine concomitantly.
In our work, the in vitro study at a pH of 7.4 did not demonstrate adsorption
of T4 to calcium carbonate. However, at a pH of 2.0, simulating
gastric acidity, there was adsorption of levothyroxine to calcium carbonate,
with 52% and 90% of 125I-T4 in the supernatant (compared
with baseline) at calcium concentrations of 400 mg/mL and 100 mg/mL, respectively.
The size of the pellet of insoluble calcium carbonate after centrifugation
did not vary between the samples at pH 7.4 and at pH 2.0.
Levothyroxine is absorbed mostly in the upper portion of the small intestine.17 Thus, adsorption at a gastric pH would only partially
explain the effect of calcium on the thyroid function studies. There may be
other mechanisms operating in the small intestine. Nonetheless, the effect
of acidity on the binding of calcium and T4 may explain why only
13 of 20 patients had a decreased free T4 level and only 4 of 20
patients had a thyrotropin level above the normal range while taking calcium.
It is possible that these patients may have increased acidity in the stomach
compared with the others. Alternatively, the patients who did not exhibit
the effect of calcium on their thyroid function tests may have had relative
achlorhydria. None of the patients were using proton pump inhibitors or histamine
H2-receptor antagonists on a regular basis. It is recommended that
calcium carbonate be taken after a meal to optimize its absorption. Thus,
the in vivo study did not exactly simulate the recommended clinical conditions.
The results of the study with 20 patients and the in vitro experiments
support the clinical practice of monitoring patients taking both calcium carbonate
and levothyroxine carefully for a change in thyroid function tests, especially
an elevated thyrotropin level. If an elevated thyrotropin level should occur,
it would be advisable to separate the time of ingestion of the calcium and