Cholesterol metabolism in hypothyroidism and

hyperthyroidism in man
Jeffrey J. Abrams and Scott M. Grundy
Department of Medicine, Veterans Administration Medical Center and
University of California, San Diego, CA 92 161

Supplementary key words low densitylipoproteins

lipids . cholesterol balance

. biliary

T h e hypercholesterolemia of hypothyroidism and

hypocholesterolemiaofhyperthyroidism
are longrecognized and well-accepted clinical findings (1-4).
The mechanismsofhypercholesterolemia
in hypothyroidism have been ascribed variously to decreased
clearance of cholesterol from plasma (5-S), reduced
conversion of cholesterol to bile acids in the liver (912), and delayed removal of low density lipoprotein
from the plasma (13). The reverse actions have been

suggested as being responsible for the low cholesterol

concentrations in hyperthyroidism.
Because of these conflicting reports, we initiated
a series of studies to examine further the actions of
thyroid hormones on the metabolism of cholesterol
and bile acids in man.These investigations were
carried out to ascertain whether lowering of plasma
cholesterol by thyroid hormones can be explained by
alterations in synthesis, excretion, or absorption of
sterols. At the same time, these same processes were
examined f o r their relation to biliary lipids to determinewhetherthyroiddysfunctionmight
modify
saturation of bile with cholesterol.

METHODS

Patients
Studies o n metabolism of cholesterol and bile
acids were performed in most of the patients with
hypo- and hyperthyroidism
described
in the accompanying paper (14), and in this paper the same
numbersare used to designate each patient. T h e
patients were divided into
three groups,nonobese and
obese hypothyroid patients,and hyperthyroid patients.
Inthe following text,thenumbers
of patients in
each study,along with theirparticularcharacteristics, will be delineated in more detail. Theseinvestigations were carried out on the
Special Diagnostic
and Treatment Unit, VeteransAdministration Medical
Center, San Diego, CA. All patients gave informed
consent for their studies.
The results from the presentsubjects are compared
to those obtainedin normal subjects who were studied
previously in our laboratory. Under each procedure
outlined below, the characteristicsof
thecontrol
subjects are described.
Abbreviations: TG, triglyceride:VLDL. very low densitylipoprotein: LDL, low density lipoprotein;HDL, highdensity lipoprotein: GLC, gas-liquid chromatography: IW, ideal weight.

Journal of Lipid ResearchVolume22,

1981

323

Downloaded from www.jlr.org by guest, on May 31, 2016

Abstract Studies were carried out on cholesterol metabolism in 11 nonobesepatientsand

16 obesepatients with
hypothyroidism and 13 with hyperthyroidism. T h e patients
underwentseveral investigations under metabolic ward
conditions. Hypothyroid patients usually had an increase
in low density lipoprotein (LDL)-cholesterol. Several mechanisms may have combined to cause a high LDL. For instance,
the obese hypothyroid patients had an increase in cholesterol
synthesis. Absorption of cholesterol also was increased
frequently.However,othermechanismsnotexplored
in
this study probably contributed to most of the fall in LDLcholesterol. Treatment of hypothyroid patients produced
the expected fall in LDL. One possible mechanism could
be thatthyroidhormonesenhancethe
conversion of
cholesterol into bile acids;thismechanism
has been suggested by otherworkersfromanimalstudies.However,
no evidence was obtained in either hypothyroid or hyperthyroid patients that thyroid hormones alter
synthesis of bile
to
acids. Ontheotherhand,thehormonesappeared
increase the synthesis of cholesterol.Patients with hypothyroidism frequently had supersaturated bile. T h e cause
of biliary cholesterol
was mostly anenhancedsecretion
associated with a tendency to obesity and increasedsynthesis
of cholesterol. In contrast, the usually thin hyperthyroid
patients did not have supersaturated bi1e.m T h e studies
show that thyroid hormones a ) influence LDL-cholesterol
by an action onthe catabolism of LDL-independent of
alterations in synthesis,catabolism, absorption, or excretion: b) stimulate synthesis of cholesterol;andc)
affect
biliary lipid metabolism in large part by influencing energy
balance and cholesterolsynthesis.-Abrams, J. J., and S . M.
Grundy. Cholesterol metabolism in hypothyroidismand
hyperthyroidism in man. j.Lipid Res. 1981. 22: 323-338.

Experimental design

o f '

324

Journal of Lipid Research

Volume 22, 19x1

Downloaded from www.jlr.org by guest, on May 31, 2016

15% of calories as milk protein, 45% a s dextrose, and

40%
as fat, mostly in the form of lard. Formulas rvere
A totalof
40 patients with thJ.roic1 disease was
prepared
by Hospital Diet ProductsChrp.,Buena
studied. Some of the patients
\\ere admitted to the
Park,
(:A.
O n e solid meal was given at 1 1 A M and it
hospital for only shortperiods(mostlyforstudies
contained
dry
cereal (corn flakes), nonfat bread, skim
oftriglyceridemetabolism)antl,except
f'or measmilk, added lard, and sugar for coffee. This diet prourements of plasma lipoproteins, the) didn o t undergo
\.itled for a l o w intake of cholesterol to f'acilitate estidetailed
investigation
of
cholesterol
metabolism.
mation o f ' cholesterol balance. Vitamin
antl minel-a1
In particular, several hypel-thyroid patients were too
supplements rvere given
daily.
Each patient was
symptomaticfromtheirdiseaseto
complete proweigheddaily
and caloricintake
\ v a s ad,justed to
longed cholesterol balance. T h e status of each paLient
maintain rveight at ;I constant lewl t h r o ~ ~ g h o u t
w i t h regard t o the period of hospitali~ation is given
the StLIdy.
in Tables 1-3 of the companion paper (14).
as deEstimations of'cholestel-ol balance ~vere made
Long-term patients, after a brief period of'stahilizascribed earlier ( 18-2 l ) . Stools were collected throughtion in the hospital, were started o n an isocaloric diet
o u t both dietary periods and usually were combined
and the intake o f ' calories vas adjusted so that body
into 4-clay pools. Fecal neutraland acidicsteroids
weightstveremaintainedconstantthroughoutthe
were isolated separately, and their masses were deterstudy. During the first month (Period I ) the patients
rvere maintained in their abnormal state (hypothyroid mined by gas-liquid chrornatograplly (GL(;) (18, 19).
G I 2 ( : analysis o f ' neutralsteroidsdistinguished
beor hyperthyroid). I n addition to the tests done on tritlveen cholester-ol and plant sterols and their steroid
glyceridemetabolismdescribed
in thecornpanion
conversion products. .\nalyses were carried out entirely
paper (14), the following ~ ~ ~ e a s ~ ~ ~ -\\.ere
c m e made:
nts
h y chemicalprocedures.
To correctfor
losses
( 1 ) plasmalipids
twice ~ v e e k l y ;1 ) ) cholcstel-ol i n the
neutral
steroids,
beta-sitosterol
was
given
a
s
capsules
differentlipoprotein
f'ractions a t the end o f ' each
(225
mg
twice
daily)
('LO),
and
excretions
o
f ' acidic
period; c ) neutraland
acidicsteroidsfromstool
steroids
were
corrected
for
variations
in
fecal
flow by
samples collected daily; d ) lipid composition o f ' gallL
I
S
o
~
l
'
chromic
oxide
(21).
l>latlder bile (determined weekly); and P ) hepatic
For control, 14 normal male subjects were studied
secretion of biliary lipids a n d pool sizes o f ' bile acids.
on
the same dietary regimen andby t h e s a ~ nanalytic
e
After appropriate treatment o f ' their disease state,
techniques.
Ages
of
controls
ranged
from
29
to
63
andreturn
t o euthyroidism,eachpatient
was re)ears
(mean
51
years);
their
Iveights
were
95115%'
studied for anothermonth(Period
11). To keep
of itteal.
patients at their pretreatnnent \\eight,
it was usually
The data on cholesterol balance are expressed
in
necessary to alter the caloric intake
a s compared t o
three ways: a ) asabsolutevalues(mg/day),
/ J ) in
Period I. Otherwise, Period I I \vas identical to Period
relationtototalbodyweight(mg/dayper
kg), and
I , and the same measurements wel-e made.
c ) i n relationtoidealbodyweight
(mg/:iday per kg
Methodology
IN'). We suggest that normalization of results to ideal
weight may in some cases provide a better cotnpa~-ison
Plasma lipid,^ cmd lipoproteins. Totalplasmachobetween groups than absolutevalues or those corrected
lesterol a n d T G weredeterminedonaTechnicon
tototalbodyweight.
One ad\:antage o f ' expressing
Auto-Analyzer(Model
11, TechniconInstruments
data in relation to ideal body weight
is that it is possible
C:orp., Tarrytown, NE' (15,16)). (:oncentrationsof
to compare groups that differ body
in size and degree
cholesterol i n very low density lipoproteins (VLDL),
of'adiposity. For example, in obese patients, if'halance
l o w densitylipoproteins(LDL),andhighdensity
data are based on actualratherthanidealweight,
lipoproteins (HDL) were estimated
a s described in the
results w i l l bedistortedtowardsinordinately
low
Lipid Research Clinics Manual of Laboratory Operavalues; in other words, dividing outputs
by a large
tions (17).
mass of adipose tissue may obscure real differences
Cholesterolbalancestudies.
Patients who underwent
in absolute production rates. This issue will be concholesterol balance studies were fed a diet of mixed
sidered in mote detail in the Discussion section.
solid food andliquid formula containing40% of calories
Lipid compositionof gallhladdu bile. Samples of fasting
as fat. This dietwas very low in cholesterol to facilitate
measurements of sterol balance;
it also contained a fat gallbladder bile were obtained three times during each
of the two study periods. Samples were analyzed f o r
content and composition similar to the normal
U.S.
cholesterol, bile acids,andphospholipids.Samples
diet. Each day the patientsweregiventhreemeals
were aspirated from a single lumen tube positioned
of liquid formula and one with solid food. Formulas
by X-ray guidance in thesecondportionofthe
were given at 8:30 A M , 4 P M , a n d 7 I'M: they contained

Abrums und Grundy

contents of formula diets (27). In

calculation of the
percent saturation of hepatic bile, we have assumed
that hepatic bile contains 3% solids (25, 26).
T h e pool sizeof bile acids was measured simultaneously
with hepatic secretion rates as described before (22).
Briefly, 5 pCi of [24-''C]cholic acid (New England
Nuclear, Boston, MA) was given intraduodenally at
thebeginning of theformulainfusion.After
4 hr
which allowed for equilibration, the ratio of isotope
to total bile acids ("specific activity") became constant
(22). A mean specific activitywas
determinedon
hourly samples over the next 6 hr, and thetotal pool
of bile acids was determined by dividing the dose of
radioactivity given by the mean specific activity.
For control, eight normal men were studied by the
same methods; their ages ranged from 29 to 60 years
(mean 47), and weights were 95- 115% of ideal.
Cholesterol absorption. Net absorption of cholesterol
was measured as recently described (28). The basic
equation for determining cholesterol absorption is as
follows:
Mass absorbed

(daily biliary secretion

+ exogenous intake

- daily fecal excretion)

Daily secretionrates of cholesterol were estimated

by multiplyinghourlysecretionrates
during continuous liquid formulainfusion by 24; this conversion is based on our recent finding that hourly secretion of biliary cholesterol x 24 closely approximates
duodenal outputs through a 24-hr period measured
in patients given three equal meals at 8 A M , 1 P M , and
6 PM (28).
Whenestimatingcholesterolabsorption
by this
method,the value obtained is thenetabsorption
of cholesterol between the upper duodenum and the
anus. If the intestine contributes cholesterol to the
lumen during the measurement, the net absorption
will be reduced correspondingly. However, as
we have
shown (28), values for percentage cholesterol absorption are as high or higher than those obtained previously by other methods; this should not
have occurred if intestinal secretion of cholesterol had been
of consequence.

RESULTS

Plasma lipids and lipoprotein-cholesterol

Data for plasma lipids in nonobese,hypothyroid
patientsbefore
andduringtherapy
with T4 are
presented in Table 1. Before treatment, total plasma
cholesterolaveraged289
& 18 mg/dl and, in most
patients, levels were reduced by T4 therapy(mean

Cholesterol metabolism in hypo- and hyperthyroidism

325

Downloaded from www.jlr.org by guest, on May 31, 2016

duodenum. Gallbladder contractionwas stimulated by

intraduodenal injection of an emulsion of safflower
oil, which \vas free of cholesterol and phospholipids.
Duodenal fluid rich in gallbladder bile was then collected by slow suction over a period of 20 min. The
collected bile (30-50 ml) was thoroughly mixed and
a 10-mlsample was retained for analysis; the remainder
was returned to thepatient via thetube. Samples
were added immediately to 30 ml ofchloroformmethanol 2: 1. Methods for separation of lipid components were the same as usedbefore in our laboratory (22).Cholesterol was measured on GLC as the
trimethylsilyl ether (19).Phospholipids were estimated
by the method of Rouser, Sidney, and Akira (23), and
bile acids were determined by a standard enzymatic
procedure(24,25).
Bile lipid composition was expressed a s molar % for each lipid component according to Admirand and Small (25). Percent saturation
was calculated by the criteria of Carey and Small (26);
these workers found thatsaturation of bile is a function
of total solids present. For calculation of percent
saturation, we have assumed that gallbladderbile contained 10% solids (25,26). For control, 14 subjects
used for cholesterol balance also were employed as
control for gallbladder bile composition. An average
of three samples was obtained from each patient.
Outputs of biliarylipids. Hourly outputs of biliary
cholesterol, bile acids, andphospholipidsduring
constant feeding of a formula diet were determined
by themarker-dilutiontechniqueofGrundyand
Metzger (27). After an overnight fast, a three-lumen
tube was positioned in the duodenum with the two
most proximal outlets adjacent to the ampulla
of Vater
andthethirdoutlet
10 cm distally. The tube was
placed in the correct position with X-ray guidance.
The same liquid formula as used throughout thestudy
was infused
continuously
throughoneproximal
lumen; P-sitosterol was also infused as adilution
marker. After allowing 4 hr for gallbladder contraction and for stabilization of hepatic bile secretion,
hourly samples were obtained
for 6 hr from the second
proximal and distal outlets by slow and continuous
aspiration. Less than 5% of intestinal contents passing
these ports was aspiratedthroughthe
tubes. Since
the inputs of p-sitosterol and exogenous cholesterol
were known with precision, the rateof hepatic cholesterol secretion could be estimated from the ratio of
cholesterol to p-sitosterol at the distal outlet. These
data combined with measurements of concentrations
of bile acids and phospholipids relative to cholesterol
at the proximal outlet permitted
calculation of the
hourly output of bile acids and phospholipids. Equations used in these calculations have been presented
previously along with correctionsforcholesterol

TABLE 1. Plasma lipids (nonobese hypothyroid patients)

~~

Total

Total

LDL

HDL

mgldl

mgldl

mgldl

332 t 44
197 f 33

234
139

308
144

59
55

226
255

184
24

1I4

11

117

123
143

64
82

I
11

152 t 13
148"
5

87
96

104
112

40
38

11

134?

159 2 20
21

104
106

95
71

36
41

302
273

150
92

142

47

Cholesterol"
Cholesterol"
Triglycerides'
Cholesterolb
Period" Patient
mgldl

1
2

3
1

1
I1
I

4
5

I1
6

-C

2
5

SD

12

2 5 0 t 16
2 1 6 2 24

145
110

189
163

56
55

1
11

242 f 38
291

131
118

200
168

38
84

I
I1

2.51
254

22
4

139
152

199
188

56
40

I
I1

303 -t 28
263 t 27

178
162

2IO
177

24
26

10

I
I1

207 t 17
144 5
7

196
130

123
89

49
35

40'

I
I1

318 t 59
2 2 5 ? 24

535
349

I59
109

42
30

Mean t SEM

I
I1

2 8 9 t 18
221 t 1 T f

183 f 37
141 ? 22

171 t 20
136 5 13/

46 t 4
49 6'

-t
-t

Period I , hypothyroid period; 11, treated period (euthyroid).

" Mean 5 SD for six determinations in last 3 weeks of each period; exceptions are Patients
6 and 8 who had only one rneasuretnent in Period 11.
' Mean value for plasma TG o f each period; for variation, see Table 4, Ref 1 1 .
" Value for o n e measurement at the end of each period.
" Patient40
was n o t included in the cotnpanion study
(14); he w a s 59 yr old and weighed
63kg(100~ofIW);hisadmissionTSHwas>100unitsandT,was~ero;theuricacidwas10.2rng/dl.
Difference between Period I and I 1 \vas significantly different at P < 0.05 by paired I-test.
" Difference between Period
I a n d 11 not statistically significant.
"

= 22 1 ? 17 mg/dl). Half of the patients had an initial

LDL-cholesterol of greater than 185 mg/dl and, in all
patientsexceptone
( N o . 3), thisfraction was decreasedafterreturn
to theeuthyroidstate.The
effects on HDL-cholesterol were variable. Some patients had an increase and others had a reduction,
with the net effect being no change in the mean.
Similarresultswereobservedforobese,hypothyroid patients (Table 2). Their degree of hypercholesterolemia was about the same as for the nonobese group. All patientsexceptonehadareduction in total cholesterol with treatment and, in most
patients, LDL-cholesterol was decreased as well. However, in those with initial hypertriglyceridemia, treatment with T4usually caused no change or an increase
in LDL-cholesterolsimultaneous
with areduction
in T G . In these obese patients, HDL-cholesterol was

326

Journal of Lipid Research

Volume 22, 1981

relativelylow,
ashasbeenreportedgenerallyfor
obese subjects (29). Treatment with T4 did not cause
a rise in HDL-cholesterol in this group.
As expected, patients with hyperthyroidism usually
had a low plasma cholesterol (Table 3). Eleven of 13
had a concentrationbelow 200 mg/dl. LDL-cholesterol
concentrations likewise were low, and both total- and
LDL-cholesterol increased with return to the euthyroid
state. Several hyperthyroid patients also had relatively
low HDL-cholesteroland,onthemean,
thisfraction increased significantly after therapy.

Cholesterol balance
Cholesterol balance data for the three groups
of
patients are shown in Tables 4, 6, and 8; based on
the assumption that these patients were
in the metabolic
steady state, their balance data were converted into

Downloaded from www.jlr.org by guest, on May 31, 2016

I1

T A B L E 2.
Paricnr

Period"

Plasma lipids(obese hypothyroidpatients)

HDL
Total
Triglycerides"

LDL
Cholesterol"

Cholesterol"

SD

mgldl

mgldl

nrgldl

Total
Cholesterol"
mgldl

1
11

221 ? 23
1 8 6 2 21

409
403

104
68

28
23

12

I
I1

297
215

260
40

199
140

46
67

13

1
11

370 4 39
370 f 17

174
195

257
208

56
50

14

I
11

256
218

200
190

158
125

60
55

15

1
I1

2 9 6 ? 31
213 ? 25

179
167

16

1
11

305 4 25
2 6 4 f 29

557
294

174
I75

32
39

261 t 26
1 7 9 4 16

137
108

149
103

61
41

11

242 f 15
1 3 9 4 30

136
105

19

I
11

541
455

103
55

1806
1177

49
56

19
18

20

I
I1

292 -+ 21
177 ? 51

292
179

176
55

23
21

377 f 78
211 t 13

870
38 1

92
98

34
31

60
40

178
166

35
40

229
166

34
34

17

I1
1

18

21

I1

11

225
2 14

23

I
11

387
274?

66
27

487
336

24

I
I1

329?
228?

32
18

178
158

I1

263 t 22
239 f 10

224
214

160
179

46
32

I
I1

256 t 23
187 4 42

179
111

181
155

48
35

I
I1

3 0 7 4 20
2 3 6 ? 20'

162 ? 15
130 r 14'

40 ? 4
37 t 4'

22

25
26
Mean

f
?

SEM

384
256

108
67'

Downloaded from www.jlr.org by guest, on May 31, 2016

I1

Period 1, hypothyroid period; 11, treated period (euthyroid).

Patients 13, 15, and 23 who had only one
measurement in each period.
Mean value for plasma TG of each period; for variation, see Table 5, Ref 14.
Value for one measurement at the end of each period.
Difference significant at P < 0.025.
'Not significant.

* Mean 5 SD for six determinations; exceptions are

values for synthesis of cholesterol and bile acids in

Tables 5 , 7, and9. Also, values forproduction of
cholesterol and bile acids are presented for 14 normal
subjects in Table 5.
Data for five nonobese,hypothyroidpatients
are
shown in Tables 4 and 5. Before therapy, cholesterol
synthesis was somewhat lower in the hypothyroid
Abrams and Grundy

state than in euthyroid controls (7.9 versus 9.6 mg/kg

perday;Table
5). Incontrast,production
of bile
acids was essentially normal before treatment. After
therapy,
four fiveof
patientsan
had
increase in outputs
of neutral steroids. While this increase was not
statistically significant for the group, total synthesis
of cholesterol was increased significantly following

Cholesterol metabolism in hypo- and hyperthyroidism

327

7'ABI.E 3. Plasmalipids (hyperthyroidpatients)

Patients

Period"

m,qldl

27
43

11

28

I1
29

31
45
32
33
37

74 114 t 19
89 156 ? 12

64
62

151 t- 25
207 t 21

46

61

86
152

116

104

46

119

29
48

I
II

94 183 t- 70
177 -c 14

103
83

1I f

56

I
II

I79 t 10
255 2 53

131
171

121
233

33

82 138 t- 17
163 t- 13

117

Ill

47

206 t 27
194
262 t 14

I24 4
130

156

129
201 t 22
309 t 25

186
23 1

210

5. 1

I
II
1

100

42

88 165
149
24 1
70 115 -c 29
89 209 -c 48

38

I
I1

117 -c 12

I
II

172 t 23
228 ? 18

1 10

I55 f 10
211 -c 13"

106 ? 10
119 t 13'

II

40

66
72

I
I1

162

21

123

37

Mean t- SEM

lll,~l~l/

55
98

II

39

lllpl~ll

? 28
179 ? 21

1
11

II
36
66

lllgl~ll

80
97

42

118
I30

1 I7
115

37

118
43

118

1.58
99 t I O
149 t- 15'

41 & 3
51 t 4'

Period I , hyperthyroid period; 11. treated period (euthyroid).

Mean f SD for six determinations; exceptions are Patient 41 who had only one measurement
in each period and Patient 38 who had one in Period 11.
Mean value for plasma T G of each period; for variation see Table 5, Ref 14.
" Value for one measurement at the end o f each period.
Differences between Periods I and I I not significant.
Differences significant at P < 0.0 1.
"

"

'

returntotheeuthyroidstate.
Despite the rise in
cholesterol synthesis, production of bile acids was not
altered by T4 therapy.
Results for seven obese, hypothyroid patients are
presented in Tables 6 a n d 7.When data were normalized to total body weight, cholesterol synthesis per kg
was only slightly increased; but when normalized to
idealweight, which is abetterreflectionofactual
production, synthesis rates for cholesterol were much
greater than normal even before treatment (Table
7).
T h e s a m e was true for bile acids.Followingtreatment with T,, fecal neutral steroids were significantly
higher, as were total steroids and Cholesterol balance
(Table 6). Bile acid excretion was unchanged. Thus,
328

Journal of Lipid Research

Volume 22, 1981

as with nonobese patients, T, therapy caused a significant rise in cholesterol synthesis, but bile acid synthesis was not altered (Table7).Thus, while treatment
with T4 caused a significant increase
in synthesis of
cholesterol, these patients were overproducing cholesterol and bile acids even before treatment.
When the data from total
a of 12 nonobese andobese
patients with hypothyroidism were combined, neutral
steroidexcretion was significantlylower
by paired
analysisbeforetreatment
(437 5 51 mg/day)than
after (621 58 mg/day) ( P < 0.05). Also, cholesterol
balance (synthesis) before treatment (848 2 117 mg/
day) was less than following therapy (IO71 5 153
mg/day) ( P < 0.05). On the other hand, combining

Downloaded from www.jlr.org by guest, on May 31, 2016

3534

SD

HIII.
Choles~c~~d"

108
101

II

34
41

1.I)I.
Ch(~le\~erol"

rrigl!ret-ide'

69 117 t 21
194 2 24

I1
30

rOral

Total
(:holestcrol"

T A B L E 4. Cholesterol balance data (nonobese hypothyroid patients)

Fecal Steroid Excretions
Acdic
Period"

Seutral

Cholesterol

Days:So.

Patient

BalanreStel-oid5
Ste~-oids

Steroids

Total

Cholesterol

1
I1

34:5
34:6

92
92

200 t 29
452 t 203

143 t 50
1 9 0 t 71

SD
343 t 66
642 t 232

251 t 66
551 t 232

I
I1

34:5
51:6

148
118

576 t 113
676 f 8 6

495 t 8 6
519 t 134

1071 ? 180
1195 f 66

923 t 140
1084 t 64

I
I1

29:5
34:6

135
135

330 t 142
4 3 9 6t 8

206 t 72
393 t 81

536 t 194
852 t 78

401 t 194
718 ? 78

I
I1

40:5
32:6

105
105

5.54 t 54
471 t 89

275 c 16
236 t 23

8 2 0 t 36
708 t 88

715 t 36
603 t 88

24:6
24:4

99
99

291 t 8 6
308 t 51

279 t 128
389 t 148

570 t 193
696 t 183

471 t 193
597 t 183

390 t 74
473 t 59"

335 f 81
345 t 59c

668 t 126
819 t 100'

552 t 119
711 t 97"

mgldn)

mgldq

10

11

Mean t SEM

116 t 1 1
110 t 8'

I
I1

Period I, hypothyroid; Period 11, euthyroid.

Duration of balance period (days) and number o f successive stool pools analyzed; the ratio
numbet- of days in each pool.
' Difference between Periods I and I1 not statistically significant.

"

"

TABLE 5.
Patienc

and bile acids. In the one of these, who was not obese
(No. 33; 113% IW), production rates of cholesterol
and bile acids were 15.9 and 10.1 mg/kg IW/day,
respectively. In the other patient,who was obese (No.
39; 144% IW), synthesis rates were respectively 21.5
and6.5
mg/kg IW/day. Twootherhyperthyroid
patients of normal weight had normal rates of syn-

Steroid synthesis data (nonubese hypothyroid patients)

Synthesis
Cholesterol
Period"

Bile Acid Synthesis

mglkglda?

rnglkg-

mglkglday

mglkgI Wldfly

I Ilildq

4
6

9
10

Mean t SEM

I
I1

4.7
10.2

4.5
9.8

2.7
3.5

2.5
3.4

I
I1

14.1
16.6

14.8
17.4

7.6
7.9

7.9
8.3

I
I1

5.2
9.3

5.7
10.3

2.7
5.1

3.0
5.6

I1

8.1
6.9

9.6
8.1

3.1
2.7

3.7
3.2

I
I1

7.2
9.2

7.4
9.3

4.3
6.0

4.4
6.1

I
I1

7.9 t 1.7
10.4 t 1.6b

8.4 t 1.8
11.0 t 1.6*

4.1 r 0.9
5.0 t 0.9'

4.3 t 1.0
5.3 t 0.9r

10.1

4.9 t 0.5

5.1 t 0.5

Normal
(14 subjects)

Mean
"

SEM

9.6 t 0.6

0.6

Period I, hypothyroid period; 11, euthyroid period.

Difference between Periods I and I1 significant at P < 0.05 by paired t tests
Difference not significant.

Abrums and Grundy

Cholesterol metabolism in hypo- and hyperthyroidism

329

Downloaded from www.jlr.org by guest, on May 31, 2016

the bile acid data failed to produce a significant difference between the two periods.
Data for hyperthyroidpatientsarepresented
in
Tables 8 and 9. Forthe small number of patients
who were able to complete the balance studies, the
results were variable. Two patients (Nos. 33 and 34)
had adistinct elevation in synthesis of both cholesterol

of the two figures gives the average

TABLE 6. Cholesterol balance data (obese hypothyroid patients)

Fecal Steroid Excretions

Pel-io&'

Patlent

Da).s:Ko.
Derernl.h

Cholestel-ol

1nr;tkr

Neutral
Srcroid5

.Acidic
Steroids

mgiday

mg/dq

583

I1

17
19

1215
I1

24:6
62: 1 2

I
I1
1

347 t 174
t 272

23:6
26:6

147
147

389 t 53
534 t 116

904 t 310
1293
850 t 1384
483

t 1237
559

1146 t 277
f 559

25:5
34:6

109

109

509
413 t 44
686 t 483
198

922 t 149
t 1169
123

t 180
t 1060
296

813 t 180
t 296

11

24:6
37:6

114

500 t 32
938 t660
191

6 5 2 t1152
27
t
1598
90

t 33
t1484
184

1041 2r 31
t 184

I
I1

33:5
35:6

862 145
160

1085
f 368
956 1644
-t 274

1947
t 288
2600
t 364

1802
t 498
2440
t 486

t 492
2

486

t 95
293
t 161

456
1044
t 74
984
f 87

899t 145
749
t 176

145

55

1120 t 58
1427 t 63

1 0 1 o5t 8
1317 t 63

12011061
2r 134
f 203"

t 135
1329 t 205"

I 588 168 34:6

60014535:6

24

11

26

1
11

31:6

29:6

I
I1

552

110
110

143
555
t 15
131 t 8"

568 t 77
3827
0 6 0t
0? 54
646
f 59
727 t 68''

t 99
734 t 165r
1474

Period 1, hypothyroid; Period 11, euthyroid.

Duration of balance period (days) and number o f ' successive stool pools analyzed; the ratio of the
number of days in each pool.
Difference between Periods I and I 1 not statistically significant
" Difference between Periods 1 and I 1 significant at P < 0.025.

929

201
714

1 I53 t 381
t 278

? 199
1018 f 380

t 176

"

"

thesis for both cholesterol and bile acids. Treatment

of three patients with hyperthyroidism produced no
change in the synthesis of either sterol.

two figures gives the average

sible in only two patients with hyperthyroidism (Table

12). These patientshad a relatively low saturation
of stimulated hepatic bile, but otherwise their values
were unremarkable.Furthermore,treatmentproBiliary lipid metabolism
duced no significant changes in any of the variables.
Data for hourly output of biliary lipids and pool
Finally, results for lipid composition of gallbladder
sizes of bile acids are presented for the three groups
bile are presented in Tables 13- 15. The gallbladder
in Tables 10-12. Table 10 compares data for four
bile of nonobese, hypothyroid patients was somewhat
higherthanthatofcomparablecontrolsubjects,
nonobese, hypothyroid patients with eight nonobese,
euthyroid subjects. T h e resultsfor the two groups
but treatment of the former did not cause a significant reduction in bile saturation (Table 23). The
were very similar for all variables. Also, in the hypothyroid patients,lipid composition and percent satura- gallbladder bile was much more saturated in obese,
hypothyroid patients which is characteristic of obese
tion of stimulatedhepatic
bile were unaltered by
patients in general (Table 14).In these patients, the
therapy.Ontheotherhand,
hepaticsecretionof
molarpercentages of thethree biliary lipids were
cholesterol was greateraftertherapy,
which is in
essentially unaffected by T4treatment, except that
two
accord with a corresponding increase in fecal neutral
steroids. The pool sizes of bile acids, which were in
obese patients (Nos. 20 and 21) had a distinct increase in bile saturation after return to the euthyroid
thenormalrange
in theuntreatedstate,
were not
state. By way of contrast, in the only two patients with
changed by treatment.
hyperthyroidism in whom the studycould be comIntheeight
obese patients with hypothyroidism
pleted, molar % cholesterol and saturation were both
(Table ll), the saturation of hepatic bile was greater
lower before treatment, butthey showed little change
than in nonobesepatients with and withouthypoin either after return
to the euthyroid state
thyroidism (Table 10);this was due to a greater secre(Table 15).
tion of cholesterol in the obese group. Treatmentwith
Cholesterol absorption
T, produced no consistent differences in either outputsofcholesterol,
bile acids, andphospholipids
Values for cholesterol absorption in eight patients
with hypothyroidism are presented alongwith results
or saturation of stimulated hepatic bile.
in ninenormal subjects in Table 16. Duringtheir
Complete studies of biliary lipid outputs were pos330

Journal of Lipid Research

Volume 22, 198 1

Downloaded from www.jlr.org by guest, on May 31, 2016

Mean t S E M

SD

135

I110

23

Cholesterol
Balance

t 175
545 t 261
608

I1
21

Total
Steroids

TABLE 7. Steroidsynthesis data (obese hypothyroid patients)

Synthesis
Cholesterol
Period"Patient

Synthesis Acid
mglkg-

mglkglday

Bile
mglkglday

mglkgIWIday

I Wlda):

11
17

8.6
12.3

10.2
14.6

4.2
7.3

5.0
8.7

13.8
14.9

18.1
19.5

10.9
10.2

14.3
13.4

9.0
11.8

12.0
15.7

5.7
5.4

7.5
7.1

10.7
15.3

15.8
22.5

6.7
6.8

9.9
10.0

14.3
19.3

22.5
30.5

8.6
13.0

13.6
20.6

8.3
6.9

13.2
11.0

4.2
2.7

6.7
4.3

8.0
10.5

20.2
26.3

4.5
4.8

11.4
12.0

10.4 ? 1.0
13.0 2 1.5*

16.0 c 1.7
20.0 f 2.6b

I1
1

19

11

21

11

23

11

24

I1

26

11
1

Mean f SEM

I1
I'

"

6.4
7.2

5
?

1.0
1.3'

9.8 2 1.3
10.9 2 2.0"

Period I , hypothyroid period; 11, euthyroid period.

Difference between Periods I and I 1 significant at P < 0.05 by paired t-test.
Difference not significant.

hypothyroidperiodthepatientsdidnothavereducedabsorption
of cholesterol (66 + 3%' versus
6 3 k 5% for normals). In fact, their percentage absorption was lower after treatment with T4.TheoreticaIly, thiscouldhavebeen
due either to a greater
flux of cholesterol into the intestine
o r to a more rapid
transit. Although biliary input was somewhat increased
by T,, it is likely that the hormone
alsostimulates
intestinal transit.

In the one hyperthyroid patient ( N o . 34) who was

studied, the results are
in accord. In the untreated
period, cholesterol input was 852 mg/day; his fecal
neutral steroids averaged 318 mgiday leaving a n absorption of 534 mg/day (63%).Following therapy, his
influx was greater, 1033 mg/day, but despite this his
fecal neutral steroids dropped to 255 mg/day. Therefore, both absolute and percentage absorption (778
mg/day and 75%') weregreaterafterreturntothe

T A B L E 8. Cholesterol balance data (hyperthyroid patients)

mglrlq

mgldq

SD

33

105

28:6

518? I l l

724

12.5

1242

103

1137

103

34

108

I
II

33:7
33:6

318

51

66

I09
I22

492
393

600
490

155

282 c 64
235 f 84

109
122

35:6

473 c 134

230

68

703

11.5

609 c 115

I1

2.
:i
5

85

380

139

259

95

639 c 224

33:5
35:6

145
14.5

I I83
600

386

161

440 f 43
87
293 c 894

33

?
I

94

II
Mean 2 SEM

97

1 (4)

113

I (3)

116

c 15

I1 ( 3 )

109

II
18'

623 ? 192
6.58 c 266
378 2 128"

419
304
262

f
f

111
50
18''

Period I , I~yperthy~.oitl; Period

11, euthyroid.
Duration o f balance period (days) and number of successive stool pools analyred; the ratio
numbe~-of days i n each pool.
Difterence between Periods I and 11 not statistirally signifirant.

.554

224

413
176

1461
749

413
176

1042 f 239
975 ? 325
674 c 118'

92.5
854
565

1623

f
f

227
305

103'

"

"

Abrums and Grundy

o f the t w o figures gives the average

Cholesterol metabolism in hypo- and hyperthyroidism

331

Downloaded from www.jlr.org by guest, on May 31, 2016

I
I1

I'ABLF. 9. Steroidsynthesis data (hyperthyroidpatients)

l',ltient

Synthesis
Cholesterol

Bile Acid S) nthesis

trlg/k,y/dq

15.9

9.0

10.1

6.6
5.2

7.5
6.0

3.8
3.I

4.3
3.6

1
I1

9.0

8.2

10.5
9.5

3.4
3.8

4.0
4.5

I
I1

14.9
7.6

21.5
11.0

4..i

6..i
4.3

1 (3)

10.2 t 2.5
7.0 5 0.9"

13.2 i-3.9
4.3
8.8 -c 1.5"

-+ 0.3
3.3 t 0.3"

11

3.5
39

I1 ( 3 )

I'

mgikg1it"//J

14. I

34

I'

u,!.$kgl~loy

33

Mean f SE41

mglkgI LVI/l//?

3.0

4.9 f 0.8
4.1 ? 0.3"

Period I , h~pertI1)roid period; 11, euthyroid period.

D i f f e ~ w ~ chetween
e
Periods I and 11 not significant.

sequences of' hypothyroidism andhyperthyroidism

on themetabolism of cholesterol and bile acids and to
determine whether abnormalities might be detected
thatcorrelate with variationsinplasmacholesterol
concentrations. At the same time, possible modifications in biliary lipids, which might be secondary
to
alteredmetabolism
of cholesterol and bile acids,
were explored.

DISCUSSION

T h e hypocholesterolemicaction
of thyroidhormones is well known,
and
hypothyroid
patients
commonly
have
elevated
plasma
cholesterol
while
lipoproteins
those with hyperthyroidismhavethereverse.
TheThepresent
workdiscloses
presentstudy was designedtoascertainthecon-tioncanaffectcholesterolconcentrations
TABLE 10.

Hourly outputs o f biliary lipids and pool sizes

Lipid Composition

Patient

mglhr

"g

7 1 0 f 116
1382 f 482

2 4 9 f 48
484 f 100

129.3
92.9

2 142
2240

41 c 5
51 c 5

1029f 159
1033 f 110

361
418

f
?

43
49

85.0
93.8

2 138
2099

17.3 ? 0.8

48 c 4

1213 f 249

415

75

86.4

4161

81.2 c 1.3
79.0 f 2.2

15.0t0.8
17.0 t 2.0

38? 5
53 f 18

1131 c 215
1365 c 548

3 2 4 f 54
450 t 175

84.1
85.5

4789
546 1

74.6 ? 1.6
75.3 c 1.0

20.2 t 1.4
18.5f 1.0

53 f 7
58 c 6

986 c 176
940 t 150

416 c 52
3 5 8 f 48

95.3
120.5

1620
1796

4.8 f 0.6
77.7 f 1.4
4.9 f 0.4* 76.6 c 0.8*

17.5? 1.1
18.5 ? 0.5*

44c 3
56 c 2'

9 6 4 f 90
1180 e 113*

338-e 35
428 ? 27*

4.7 c 0.4

47
15.6 f 2.1

1139 ? 180

340 c 46

2.0

17.6 f 1.5

6.2 f 0.8
4.7 -t 1.4

76.7
76.5

0.8
7.5

17.2f 1.1
18.8 f 6.2

45
60

8
11

4.1 c 0.7
4.9 f 0.6

78.1 c 2.3
75.5 ? 1.1

17.7 -e 1.6
19.6 f 0.8

4.1 c 0.6

78.5

1.1

3.6 f 0.5
4.0 f 0.3

5.1 f 0.7
6.1 c 0.4

Normal
8 patients

SD"

85.2

78.0

Mean e SEM

-C

.Acid
Pool Size

41

0.7

79.6

1.0

f
f

0.7

9 4 0 f 204

326f

98.4 f 10.6 2672

98.2 c 7.7b 2899
106.3 f 3.2

T h e data include means f SD for six hourly samples during the steady state period of formula infusion.
statistically significant between Periods I and 11.
Difference significant at P < 0.025.

* Difference not

332

Bile
Saturation
Bile
(3%solids)

1928
2370

335 8
37 f 4

4.1

Biliary Lipid Output

f SD"

I'

of bile acids (nonobese, hypothyroid patients)

Period
Cholestet-ol
Phospholipids
Acids
Bile Chole~terol
Phospholipids
Acids
Bile
mhr

thatthyroiddysfuncin each of

Journal of Lipid Research Volume 22, 1981

2921

-t

716

f 859*

434

Downloaded from www.jlr.org by guest, on May 31, 2016

euthyroid state. Again, it seems likely that the more

rapid intestinal transit associated with hyperthyroidism
had reduced cholesterol absorption.

TABLE 11.

Hourly outputs of biliary lipids and pool sizes of bile acids (obese, hypothyroid patients)
Lipid Composition

Patient
Period
Cholesterol

Bile Acids

Biliary Lipid Output

Phospholipids
Cholesterol

Bile Acids

molar % k SD"

mgihr

Phospholipids

Bile
Saturation
(3% wlida)

Bile .\cid
Pool Size

Nlg

* SD"

5.1
7.1

f 0.6
f 0.9

77.5 f 1.5
70.7 f 2.5

17.2 1.2
22.0 f 2.6

52 f 11
78 ? 12

976 + 213
1002 + 199

335 t 90
515 f 106

107.0
123.0

3663
2.560

16

I
I1

4.8 + 0.6
5.9 f 0.9

81.1 + 4.7
70.5 t 4.9

14.0 + 4.5
23.4 f 5.4

63 + 20
42+ 7

1992 f 707
663 + 135

600 + 307
348 t 124

116.7
99.0

2959
2.569

17

I
I1

5.1
6.1

1.0
1.0

78.9 + 1.5
73.8 + 3.5

15.9 + 1.1
20.0 + 2.6

58+ 5
742 8

1190 f 189
1238 + 188

363 2 51
501 -+ 57

113.3
1 13.9

4670
3755

19

I
I1

3.7 f 0.3
5.4 f 0.5

72.9
73.1

+
+

1.5
1.5

43+ 8
582 5

1089 f 226
1017 2 188

544 + 103
463 t 49

63.2
96.6

4826
34.53

20

I
I1

5.8
4.8

+ 1.7
+ 0.5

78.1
77.4

+ 3.3

16.0 + 2.1
17.5 c 1.2

16t 9
405 7

1 0 5 f 65
2 9 0 t 45

127.6
99.6

2073
1861

21

I
I1

8.4 f 1.6
4.5 f 1.0

73.6
74.9

+ 3.3

17.9 + 3.0
20.5 f 1.2

71 f 15
72% 4

+ 169
+ 112
844 + 254
1686 + 417

315 t 85
716 + 170

167.0
83.6

4329
3824

23

I
I1

8.0
9.8

72.9

17.9 + 0.9
17.0 f 1.7

86+ 4
78 + 20

1030 + 113
775 2 274

384
297

1.6

25
68

159.5
199.7

346 1
2064

I
I1

8.6 f 1.5
7.5 + 0.5

+ 0.9

19.3 f 1.0
18.3 f 0.7

77+ 6
75 + 13

862 146
971 t 223

360
373

+ 66

74.1

83

161.7
147.8

2591
1413

I
I1

6.2
6.4

+ 0.7
+ 0.6'

76.0
73.4

17.7 + 1.0
20.0 2 0.8*

58
65

1038 + 165
1021 2 114b

376
438

f
f

53
50"

26
Mean

SEM

1.5

+ 1.6
74.0 + 1.4

f 0.6

1.8
1.9

+ 3.1
72.1 + 1.5
1.2

f 0.8b

23.3
21.4

IT

8
6'

320
815

-t

127.0 2 12.4 3572

122.1 + 13.gb 3085

+ 353

+ 446'

The data include means r SD for six hourly samples during the steady state period of formula infusion.
1 and 11.
Difference significant at P < 0.025.

' Difference not statistically significant between Periods

the different lipoprotein fractions. The major alteration occurred in the LDL fraction. In both nonobese
and obese
patients
with hypothyroidism, LDLcholesterol was relatively high and treatment with T,
produceda
20% reduction in both groups. Conversely, return to theeuthyroidstate
in patients
with hyperthyroidism evoked a mean rise in this fraction of 51%.
The influence of thyroiddysfunction
on HDLcholesterol seems more complex. Our hyperthyroid
patients had lower levels before treatment than after.
Thus, excess circulating thyroid hormone apparently
TABLE 12.

lowers HDL as it does LDL. A similar result has

been reported beforehand by Sachs, Wolfman, and
Murthy (30), and more
recently by Scottolini et al. (31).
Of interest, several of our hypothyroid patients also
had relatively low values of' HDL-cholesterol before
treatment.However,decreased
values were mostly
present in the obese groupand several nonobese
patients actually had relatively high levels, as also was
reported by Scottolini et al. (31). Both Sachs et al.
(30) and Scottolini et al. (31)treatedhypothyroid
patients with thyroid hormone and noted a depression
in HDL-cholesterol. This response was not reproduced

Hourly outputs of biliary lipids and pool sizes of bile acids (hyperthyroid patients)
Lipid Composition

Patient

Period
Cholesterol

Bile Acids

Biliary Lipid Output

Phospholipids
Cholesterol

Bile Acids

molar %"

33
34

I
I1
I
I1

Mean

+ SEM

I
I1

3.4
5.0

+ 0.6

mglhr

20.7
21.4

3.6 f 0.6
4.6 e 0.9

79.1
77.4

17.2 + 2.9
17.9 + 1.1

31
39

3 . 5 e 0.1
4.8 c O.Zb

77.4 t 1.7
75.5 c l.gb

19.Of 1.8
19.7 f 1.8b

4 8 f 17
45 f 6 b

2.9
1.6

f 4.8
f 1.8

64 f 12
51 c 1 1

75.7 t 5.0
73.5 c 2.3

f 0.6

+
+

4
11

Phospholipids

Bile
saturation
(3% solids)

Bile Acid
Pool Size

mR

SD"

1804 f 270
966 c 207

774 -t 215
435 c 84

63.2
89.7

2832
55 16

+ 190
+ 394

307 + 85
357 -c 140

76.4
94.2

2 167
2037

541 t 234
396 + 3gb

6 9 . 8 + 6.6
92.0 c 2.2'

2 5 0 0 c 333
3777 f 1740b

902
998

907 f 897
982 t 16'

The data include means t SD for six hourly samples during the steady state period of formula infusion.
Difference not statistically significant between Periods I and 11.

Abrums and Grundy

Cholesterol metabolism in hypo- and hyperthyroidism

333

Downloaded from www.jlr.org by guest, on May 31, 2016

I1

15

T A B L E 13.

Lipid composition of gallbladder bile and saturation indices

(uonobese hypothyroid patients)
%

Lipid Composition"
Patient

Pel t

o d

Chohter-ol

Bile i c i d s

Phospholipids

Satur-ation
(I0% solid$)

nroir R

7.6

71.4

21.0

109

I
I1

12.3
9.5

71.0
69.0

16.7
22.0

198
131

11

7.7
6.9

71.8
I 3.3

20.4
1i . 6

112
111

10.4

74.3

19.6

1.57

I
11

8.2
11.2

72.2
66. I

19.6
22.5

122
150

I (3)
1 (3)
11 ( 3 )

9.2 t 0.4
10.1 2 0.8
9.2 f 0.7"

72.1 t 0.3
71.7 t 0.2
70.1 2 1.6"

19..5

2 0.3
18.9 ? 0.6
20.7 2 0.3"

139 t 8
144 t 16
1 3 1 t 7"

7.7 t 0.4

74.8 t 0.9

18.6 t 0.6

120t 4

Mean 2 SEM

Normal- 14
subjects
Slean 2 SEM

Values represent mean o f three to four determinalions it1 each period.

Differences between Periods I and I1 not statistically significant.

in out patients, perhaps because

of their relatively
l o w le\~elsbeforetherapy. ,4t least t w o factors may
havebeenresponsiblefor
low HDLinourobese,
hypothyroid patients. First, obesity itself is accompanied
by reducedHDL (20), andsecond,manyofthese
patients had hypertriglyceridemia which is known to
be associated with a decreased HDL-cholesterol ( 3 2 ) .
Furthermore, with therapy, the HDL-lowering action
of T4 may ha\.e been offset to some extent h y a fall
in plasma TG.

Cholesterol balance
T h e decline of LDL-cholesterol induced by thyroid
hormones might he explained by modification of the
metabolism of cholesterol o r bile acids. Two processes
havebeenproposed
by otherworkers (9-12, 33):
thyroid hormones may a ) stimulate conversionof cholesterol into bile acids or h ) accentuate outputsof neutral steroids. Conceivably, as suggested by the current
study, cholesterol absorption may also be depressed

'I'AULE 14. Lipid composition otgallblatlder bile and satul-atiotl indices (obese hpothyroid patients)
R

Lipid Composition"
Patient

Pet iod

(:holcter-ol

Bile 4ctcls
rNOIY

15

18
I9
20
1

I
I1

334

69.3
66.6

18.9
20.9

174
172

6.9

6. 1

75.0
73.8

18.6
10.0

108
91

I
II

7.7
9.2

67.7
69.0

24.6
2 1.6

100
128

I
11

7.8
12.6

73.5
66.1

18.8
21.2

120
173

x. 1
21.4

71.0
65.5

20.8

11

13.0

116
378

11

13.2
12.8

72.2
72.8

14.5
14.3

232
228

19.4 -+ 1.3
16.8 i 2.1

142 4 21
194 t 41

I
11

I'

23
Mean 2 SEM

1?.4

11

21

11.7

Phospholipids

Saturation
(101, solids)

9.2 r 1.0
12.4 f 2.1

71.3
69.0

2
4

1.1
1.5

Values represeut meau of three t o fout- determinations in each prriod.

Journal of Lipid Research Volume 22, 198 1

Downloaded from www.jlr.org by guest, on May 31, 2016

I'

I'

"

TABLE 15. Lipid composition of gallbladder bile and saturation indices (hyperthyroid patients)
~~~~

Composition
Patient

Cholesterol
Period

Bile Acids

Phospholipids

ro

molp %

Mean

~~

%
Saturation
(10% solids)

Lipid

33

I
I1

3.4
5.0

75.7
73.5

20.7
21.4

50
72

34

I
I1

3.6
4.6

79.1
77.4

17.2
17.9

60
75

77.4 ? 1.7
75.5 2 1.9

19.0 2 1.8
19.7 2 1.8

SEM

11

3.5
4.8

f 0.1
?

0.2

55
73

5
1

TABLE 16. Cholesterol absorption(hypothyroidpatients)

~~

~~

~~

Cholesterol Input
Dietar)

Period

Cholesterol
Absorption
Biliary Cholesterol Output

Patient

mglday

mglday

mglday

792
888

200
452

684
528

77
54

I1
4

I
I1

148
118

1080
1440

576
676

652
882

53
57

I
I1

135
135

984
1224

330
459

789
900

71
66

I
11

105
105

1272
1392

554
47 1

823
1026

60
69

19

I
I1

109
109

1032
1392

413
686

728
815

64
54

21

I
I1

110
114

1704
1728

500
938

1314
904

72
49

23

I
I1

145
160

2064
1872

862
956

1347
1076

61
53

26

I
I1

110
110

1848
1800

552
827

1406
1087

72
57

119f 7
118 f 7

1347 f 164
1473 2 118

484 f 70
683 f 74

96 f 7

1063 f 97

403 f 39

Mean

mglday

92
92

SEM

I
I1

967
902

757f
~~~

Abrams and Grundy

?
f

116
64

58

66 -+ 3
57 -c 3

65

f5

Cholesterol metabolism in hypo- and hyperthyroidism

335

Downloaded from www.jlr.org by guest, on May 31, 2016

need not be the cause of plasma cholesterol lowering.

Raised steroid
outputs
could
result
either
from
mobilization of cholesterol from existing pools (as
plasma) or from promotion of cholesterol synthesis.
The former should cause a temporary rise in steroid
outputs, the latter a sustained increment. Examination of individual data favored the latter, i.e., a sustained elevation. Therefore, thyroid hormones probably stimulate cholesterol synthesis, and it seems unlikely that this action is related to their facility to
lower LDL.
Several preceding workers have reportedthat
cholesterol synthesis is suppressed in hypothyroidism

to some extent by the hormones. The results of our

research corroborate prior reports that thyroid hormonesincrease outputs of neutral steroids. In our
hypothyroid patients, these outputs
were greater after
T4therapy than before. Those with hyperthyroidism
had the reverse: their neutral steroid excretion
was
higher before treatment than after.
By way of contrast, outputs of' acidic steroidswerenotmodified
by treatment in either group. This
implies that thyroid
hormones do not augment conversion of cholesterol
to bile acids in man.
Although the excretion of neutral steroids is seemingly enhanced by the thyroid hormones, this action

Downloaded from www.jlr.org by guest, on May 31, 2016

andincreasedinhyperthyroidism
(5, 34-41).Our
Biliary lipid metabolism
data are in agreementwith these reports in a relative
T h e influence of hypo-andhyperthyroidismon
but not in an absolute sense. The net effect of treatbiliary lipid metabolism alsowas explored in this study.
ment of hypothyroid patients with
T, was to augment Preceding work in rats indicated that cholesterol consignificantly the balance of cholesterol, and hence to
centrations in bile fall after these animals are made
stimulate synthesis. From these findingswe must conhypothyroidand risewhenthey
arehyperthyroid
clude that a lack of T, slackens production of choles(10,45,46). These findings were not confirmed in our
terol. Nonetheless, obese, hypothyroid subjects regustudies in man. Two hyperthyroid patients did not
larly had elevated synthetic rates. This phenomenon
have increased cholesterol secretion, and several paprobably was due in part to an
excessive intake of
tients with hypothyroidism did. An increased biliary
calories required to maintain their obesity; the overcholesterol in obese, hypothyroid patientsis in accord
productionofcholesterol
in obesepatients is wellwith our previous findings in obese, euthyroid subrecognized (42). An additional factor that may have
jects (42). This high secretion
of biliary cholesterol
contributedto excessive synthesisofcholesterol
in
in our hypothyroid patients evidently
was the cause of
thesepatients is adiminishedoxidationofcaloric
the supersaturation of gallbladder bile often found
substrate. As Keyes and Heimberg (43) observed for
in these patients. Furthermore, the failure to correct
other lipids, the hypothyroid state
may divert 2-carbon this abnormality completelyby treatment with T4subfragments away fromoxidationtocholesterolsynstantiates the concept that the state of nutrition (i.e.,
thesis. Thus, in our obese patients, the mean produc- obesity o r nonobesity) was thecrucialdeterminant
tion of cholesterolin the hypothyroid statewas greater
of bile saturation.Thus,theeffectsofthyroid
than normal (16.0 k 1.7mg/day/kg I W versus10.1
hormones on biliary lipids must be due in large part
mg/day/kg I W for normals). Nevertheless, it was ento their influence on caloric balance.
hanced even more after treatment (to 20.0t 2.6 mg/
day/kg IW). This finding again reveals that T 4 can
Obesity and hypothyroidism
increase synthesis. Nonetheless, we conclude that the
T h e division of our hypothyroid patients into noncaloricbalance,asreflected
by obesity ornormal
obese and obese patients has been revealing. The nonweight, is more important for controlling the level of
obese patients clearly had hypercholesterolemia that
cholesterol synthesis in hypothyroidism than are con- could not be explained by overproduction of cholescentrations of circulating T,. This conclusion also is
terol or decreasedsynthesis of bile acids.This strongly
supported by the four balance studies donein hypersuggests that other factors, such as decreased catabthyroid patients (Table 9).
olism ofLDL,are
mainly responsibleforelevated
To summarize, alterations in synthesis, catabolism,
LDL-cholesterol. On the other hand,
a majority o f o u r
absorption,orexcretion
of cholesterolcannotexpatients with hypothyroidism were obese, and weight
plainfully the changes in LDL-cholesterol found in
gain has been reported to be a presenting feature
in 50
patients with thyroid dysfunction. They nevertheless
t o 75% ofpatients
with hypothyroidism(47-49).
could play a supporting role. For example, in the unPresumably, the increase in weight is due to a reductreated, hypothyroid state, many patients have an intion in oxidative utilization of calories. Once
obesity
crease in synthesis o f cholesterolandsecretion
of
is established and maintained, it can become an inVLDI,.
Both
could
contribute
t o elevated LDLdependent f-actor regulating lipid metabolism. Thus,
cholesterol.Furthermore,absorption
was relatively
despite their hypothyroidism, our obese patients exhigh in hypothyroid subjects which should bolster an
hibited an abnormally high synthesis
of cholesterol
elevated LDI,. However, these factors probably are notthat may have contributed in part
to hypercholesterolcrucial. More likely, LDL-cholesterol is influenced to
emia. O n the other hand, by keepingbodyweight
a greater extent by an independent action of thyroid
constant in our obese subjects, we were able to show
hormones on the catabolism of LDL. For example,
that thyroid hormone, independent of body weight,
Waltonet al. (13)havereported
a diminishedrecan stimulate the synthesis of cholestero1.a
moval of '"I-labeled LDL i n hypothyroidpatients
This work was supported in part by the Veterans Adminisandaccentuatedclearance
in those with hyperthytration: Dr. JeffreyJ. Abrams was an Associate Investigator
roidism. A stimulation of LDL catabolism by thyroid
and Dr. Scott M. Grundy is a Medical Investigator of the
hormones also is supported by the preliminary report Veterans Administration. The investigation was also supof Chait, Albers, and Bierman (44);they observed that ported by Grant AM-I6667 from the National Institute o f
'r4 enhances both binding and degradation of LDL Arthritis, Metabolism, and Digestive Diseases a n d No. HLin cultured human fibroblasts.
14197 awarded by the National Heart, Lung, and Blood

336

Journal of Lipid Research Volume 22, 1981

Abrams and Grundy

Cholesterol metabolism in hypo- and hyperthyroidism

337

Downloaded from www.jlr.org by guest, on May 31, 2016

Institute,HDS/DHHS. Theauthors wish to expresstheir

teins in hypothyroidismand thyrotoxicosis. Clin.
Sci.
appreciation
Marjorie
to Whelan,
Joan
Rupp,
Lianne
29: 217-238.
Leipper, and others of the Nursing and
Dietetic Services
14. Abrams, J. J., and S . M. Grundy. Metabolism of plasma
o f t h e VeteransAdministration Medical Center, SanDiego.triglycerides
in hypothyroidismandhyperthyroidism
in man. J. Lipid Res. 22: 307-322.
Excellent technical help was also provided by Robert
15. Block, W. D., K. J. Jarrett,and J. B. Levine.1965.
Ronimus, J a n n a Naylor, James Hobza, Lynne Lesh, Susan
Use o f a singlecolor reagent to improvetheautoButler, Richat-d Earl, and Leanna Johansen.
In
mateddetermination of serum totalcholesterol.
Manumi>t rrctwrd 29.jnnualy I980 and in revised fonn 12 September 1980.
Automation in Analytical Chemistrv. MediadInc..
New York. 345-347.
16. Kessler-, G., and H. Lederer. 1965. Fluorometric measurements of triglycerides. In Automation in Analytical
Chemistry. Mediad Inc., New York. 341-344.
REFERENCES
17. Lipid Research Clinics Program Manual ot Laboratory
Operations. 1974. DepartmentofHealth,
Education
1. Epstein, .4. A,, and H. Landc. 1922. Studies o n blood
and Welfare Publication NIH 75-628.
lipoids. I. T h e relation of cholesterol and protein de18. Grundy, S. M., E. H. Ahrens, Jr., and T.A. Miettinen.
Med. 30:
ficiency to basal metabolism. Arch.Intern.
1965. Quantitative isolation and gas-liquid chromato563-577.
graphic analysis of total fecal bile acids. J. LipidRes.
2. Mason, R. L . , H. M. Hunt, andL. Hurxthal. 1930. Blood
6: 397-410.
cholesterol values in hyperthyroidismandhypothy19. Miettinen, T. A., E. H. Ahrens, Jr., and S. M . Grundy.
roidism-theirsignificance.
N . Engl. J. Med. 203:
1965. Quantitative isolation and gas-liquid chromato1273- 1278.
graphic analysis of total dietaryand
fecal neutral
3. Peters, J. P., and E. B. Man. 1943. T h e interrelations
steroids. J. Lipid Res. 6: 4 11-424.
of serum lipids in patients with thyroid disease. J. Clin.
20. Grundy, S. M., E. H. Ahrens, Jr., and G. Salen. 1968.
Invrst. 22: 715-720.
Dietary beta-sitosterol as an internal standard to correct
4. Peters, J. P., and E. 8 . Man. 1950. T h e significance of
for cholesterol losses in sterol balance studies. J. Lipid
serumcholesterol in thyroid disease. J . Clin. Invest.
Res. 9: 374-387.
29: 1-11.
21. Davignon, J., W. J. Simmons, and E. H. Ahrens, Jr.1968.
5. Kurland, G. S., and J. L. Lucas. 1955. T h e metabolism
Usefulness of chromic oxide as an internal standard
of intravenously infused CL4-labeled cholesterol
in man
for balance studies in formula-fedpatientsandfor
before and after the induction of myxedema. J. Clin.
assessment
of caloric function. J. Clin.Inr~est. 47:
Invrst. 34: 947.
127- 138.
6. Kurland, G. S., J. L. Lucas, and A. S. Freedberg. 1961.
22. Grundy, S. M. 1975. Effects of polyunsaturated fats on
T h e metabolism of intravenously infused C14-labeled
lipid metabolism in patients with hypertriglyceridemia.
cholesterol in euthyroidism andmyxedema.J. Lab. Clzn.
J. Clin. Invest. 55: 269-282.
Mrd. 57: 574-585.
23. Rouser, G., F. Sidney, and Y. Akira. 1970. Two dimen7. Chobanian, A. V., B. A. Burrows, and W. Hollander.
sional thin-layer chromatographic separation of polar
1962. Body cholesterol metabolism in
man. 11. Measurelipids and determinations of phosphorus analysis of
ment of the
body cholesterol miscible pool and turnover
spots. Lipidi. 5 : 494-496.
rate. J. Clin. Inve.it. 41: 1738-1744.
8. Rabinowitz,,J. L., T. Rodman, andR. M. Myerson. 1963.
24. Talalay, P. 1960.Enzymatic analysis of' steroidhorEffect ofdextrothyroxine on metabolism of CI4-labeled
mones. Methodr Biochem. Anal. 8: 119- 143.
cholesterolandtripalmitin.
J. Am.Med. Assoc. 183:
25. Admirand, W. H., and D. M. Small. 1968. The physico758-760.
chemical basis of cholesterolgallstone
formation in
9. Friedman, M., S . 0. Byers, and R. H. Rosenman. 1952.
man. J. Clin. Inuest. 47: 1043- 1052.
Changes in excretion ofintestinal cholesterol and sterol
26.Carey, M. C., and D. M. Small.1978. T h e physical
digitonides in hyper- and hypothyroidism. Circulation.
chemistry of cholesterol solubility in bile. Relationship
5: 657-660.
to gallstone formation and dissolution in man. J. Ctin.
10. Rosenman, R. H., S. 0. Byers, and M. Friedman. 1952.
Invest. 61: 998- 1026.
T h e mechanism responsible for the altered blood cho27. Grundy, S . M., and A. L. Metzger. 1972. A physiologic
lesterol content in derangedthyroid states. J. Clin.
method for estimation of hepatic secretion of biliary
Endocrinol. Metab. 12: 1287- 1299.
lipids in man. Gastroenterology. 62: 1200-1217.
11. Ericksson, S. 1957. Influence of thyroid activity on ex28. Mok, H. Y. I., K. von Bergmann, and S . M. Grundy.
cretion of bile acids and cholesterol in the rat. Proc.
1979. Effects of continuous and intermittent feeding
SOL.Exp. Biol. Med. 94: 582-584.
on biliary lipid outputs in man: application for measureD12. Strand, 0. 1962. Influence of propylthioracil and
ments of intestinal absorption of cholesterol and bile
and L-triiodothyronine on excretion
of bile acids in bile
acids. J . Lipid Res. 20: 389-398.
fistula rats. Bile acids and steroids 118. Proc. SOC. Exp.
Biol. Med. 109: 668-672.
29. Gordon,T., W. P. Castelli, M. C. Hjortland, W. B.
13. Walton, K. W., P. J. Scott, P. W. Dykes, and J. W. L.
Kannel,and T. R. Dawber.1977. High density lipoDavids. 1965. T h e significance of alterations in serum
protein as a protective factor against coronary heart
disease. Am. J. Med. 62: 707-714.
lipids in thyroiddysfunction. 11. Alterationsofthe
metabolism and turnover of 1311-low-densitylipopro30. Sachs, B. A., L. Wolfman, and G. Murthy. 1968. Lipid

338

Journal of Lipid Research Volume 22, 198 1

40. Lipsky, S. R., P. K. Bondy, E. B. Man, and J. S. McGuire,

Jr. 1955. T h e effects of triiodothyronine on the biosynthesis of plasma lipids from acetate-l-C14 in myxedematous subjects. J . Clin. Invest. 34: 950.
41. LeRoy, G. V. 1956. Studies of cholesterol synthesis in
man using carb~n'~-labeledacetate. Ann.Int.Med.
44: 524-533.
42.Bennion, L. J., and S. M. Grundy. 1975.Effects of
obesity and caloric intake on biliary lipid metabolism
in man.]. Clin. Invest. 56: 996- 101 1.
43. Keyes, W. G.,and M. Heimberg. 1979. Influence of
thyroid status on lipid metabolism in the perfused rat
1iver.J. Clin. Invrst. 64: 182- 190.
44. Chair, A., J. J. Albers, and E. L. Bierman. 1978. Regulation of low density lipoprotein degradation by triiodothyronine. Clin. Res. 26: 303A.
45. Van Zyl, A. 1957.Note on the effects of thyroidectomy and thyroid hormone administration on the concentration of bile cholesterol and cholic acid. J . Endocrinol. 16: 213-216.
46. Thompson, J. C., and H. M. Vars. 1953. Biliary excretion of cholic acid and cholesterol in hyper-, hypo-,
andeuthyroid
rats. Proc. Sac. Exp. Biol.
Med.
83:
246-248.
47. DeCroot, L. J. and J. B. Stanbury. 1963. Adult hypothyroid states. In The Thyroid and Its Disease. J. H.
Means, editor. McGraw-Hill, New York. 287-353.
48. Watanakunakorn, C., R. E. Hodges, and T. C. Evans.
1965. Myxedema. Arch. Intern. Med. 116: 183- 190.
49. Robbins, J.,J. E. Rall, and P. Gorden. 1974. The thyroid
and iodine metabolism. In Duncan's Diseases of Metabolism. P. K. Bondy and I,. E. Rosenberg,editors.
M'. B. Sanders Co., Philadelphia. 1004-1009.

Downloaded from www.jlr.org by guest, on May 31, 2016

and clinical response to a new thyroid hormone combination. Am. J . Med. Sci. 256: 232-238.
31. Scottolini, A. G., N. V. Bhagavan, T. H. Oshiro, and
S. Y. Abe. 1980. Serum high-density lipoprotein cholesterol concentrations in hypo- and hyperthyroidism.
Clan. Chem. 26: 584-587.
32. Myers, L. H., N . R. Phillips, and R. J. Havel.1976.
Mathematical evaluation of methods for estimation of
the concentration of the major
lipid components o f
human serum lipoproteins. J . Lab. Clin. Med. 88: 491505.
33. Miettinen, T. A. 1968. Mechanism of serum cholesterol
reduction by thyroid hormones in hypothyroidism. J .
Lab. Clin. Med. 4: 537-547.
34. Byers, S. O., R. H. Rosenman, M. Friedman, and M. W.
Biggs. 1952.Rate of cholesterolsynthesis in hypoand hyperthyroid rats. J . Exp. Med. 96: 513-516.
35. Rosenman, R. H., M. Friedman, and S. 0. Byers. 195 1.
Changes in biliary cholesterol in abnormalthyroid
states. Science. 114: 210-211.
36. Marx, W., S. T. Gustin, and C. Levi. 1953. Effects of
thyroxine, thyroidectomy, and lowered environmental
temperature on incorporation of deuterium into cholesterol. Proc. SOL. Exp. Biol. Med. 83: 143-146.
37. Karp, A., and D. Stetten, Jr., 1949. The effect of thyroid
activity on certain anabolic processes studied with the
aid of deuterium. J . Biol. Chem. 179: 819-830.
38. Gould, R. G., G.V. LeRoy, G. T. Okita, J. J. Kabara,
P. Keegan,and D. M. Bergenstal.1955. The use of
C'4-labeled acetatetostudy
cholesterol metabolism
in man. J . Lab. Clin. Med. 46: 372-384.
39. Gould, R. G. 1959. T h e relationship between thyroid
hormones and cholesterol biosynthesis and turnover.
In Hormones and Atherosclerosis. G. Pincus, editor.
Academic Press, Inc., New York. 75-88.