hypothyroidism: a systematic review and meta-analysis

©The Japan Endocrine Society

2020, 67 (7), 719-732

Original

Levothyroxine in the treatment of overt or subclinicalhypothyroidism: a systematic review and meta-analysis

Yan Chen and Hai-Yan Tai

Department of Endocrinology, Wuwei People’s Hospital, Wuwei, Gansu Province, China

Abstract. The goal of this study was to review relevant randomized controlled trials in order to determine the clinicalefficacy of levothyroxine in the treatment of overt or subclinical hypothyroidism. Using appropriate keywords, we identifiedrelevant studies using PubMed, the Cochrane library, and Embase. Key pertinent sources in the literature were also reviewed,and all articles published through December 2019 were considered for inclusion. For each study, we assessed odds ratios(ORs), mean difference (MD), and 95% confidence interval (95%CI) to assess and synthesize outcomes. We included 25studies with totally 1,735 patients in the meta-analysis. In the patients with hypothyroidism, compared with L-T4, L-T4 plusL-T3 significantly decreased TSH levels and increased FT3 levels. Compared with placebo, L-T4 significantly increased FT4levels and decreased TSH levels. In patients with subclinical hypothyroidism, compared with placebo, L-T4 significantlydecreased SBP, TSH, T3 and TC and increased FT3 and FT4.

Key words: Levothyroxine, Hypothyroidism, Meta-analysis

THYROID HORMONE is an important hormoneinvolved in human lipid synthesis and energy metabo‐lism. Hypothyroidism can be caused by lesions in thehypothalamic-pituitary-thyroid axis, leading to insuffi‐cient thyroid hormone secretion. It can be divided intoprimary hypothyroidism, central hypothyroidism, andthyrotrophin or thyroid hormone insensitivity syndromeaccording to the site of the lesion, with primary hypo‐thyroidism having the highest incidence. The prevalenceof hypothyroidism is related to TSH diagnostic cut-offpoint, age, sex, race and other factors. Hypothyroidism isparticularly common in women and the elderly and ischaracterized by inadequate action of thyroid hormonesin the target tissue. The prevalence of hypothyroidism isabout 5%–10%, and the prevalence of subclinical hypo‐thyroidism is higher than that of clinical hypothyroidism.The National Health and Nutrition Survey (NHANESIII) of people aged >12 years in the general populationshowed that using 4.5 mIU/L as the TSH normal limit,the prevalence of subclinical hypothyroidism was 4.3%,and clinical hypothyroidism prevalence was 0.3% [1-3].Since thyroid hormone is an important hormone regulat‐

Submitted Dec. 13, 2019; Accepted Mar. 2, 2020 as EJ19-0583Released online in J-STAGE as advance publication Mar. 31, 2020Correspondence to: Hai-Yan Tai, Department of Endocrinology,Wuwei People’s Hospital, No. 24 North Xuanwu Street, LiangzhouDistrict, Wuwei, Gansu Province, 733000, China.E-mail: [email protected]

ing human energy metabolism, hypothyroidism can leadto insufficient energy supply and decreased basic metab‐olism in patients, resulting in chills, decreased cardiacoutput and other clinical manifestations. At the sametime, studies have shown that thyroid hormone canautonomously regulate the whole human lipid metabolicprocess and absorb serum free fatty acids and promotelipid decomposition, and hypothyroidism patients due tolack of thyroid hormone combined with hyperlipidemia.At present, preparations of thyroxine include levothyrox‐ine, triiodothyronine and its compound preparations, andthyroid tablets of animal origin, etc [4-6].

The aim of this study was to perform a meta-analysisof all available literature to obtain updated evidence toexplore the clinical efficacy of levothyroxine in the treat‐ment of overt or subclinical hypothyroidism, and to pro‐vide a basis for the selection of clinical treatment.

Material and Methods

Search strategyTo identify studies on treatment of overt or subclinical

hypothyroidism, we reviewed the Cochrane library,PubMed, and Embase databases for relevant articlespublished through December 2019. We also reviewedthe references of all identified articles to identify addi‐tional studies. Search terms were as follows: levothyrox‐ine, levothyroxine sodium, sodium levothyroxine tablet,

L-thyroxine, hypothyroidism, hypothyrea, thyroid hypo‐function, primary hypothyroidism, subclinical hypothy‐roidism, random, randomized controlled trials, and RCT.These terms were used in combination with “AND” or“OR”. This literature review was performed independ‐ently by two investigators, with a third resolving any dis‐putes as needed.

Following the PICOS (Participants, Interventions,Comparisons, Outcomes and Study design) principle,the key search terms included (P) adult patients withhypothyroidism; (I) in the treatment group, patients weretreated by levothyroxine or combined with other drug;(C/O) the outcomes including the related clinicalindexes; (S) RTC.

Study selection criteriaIncluded studies met the following criteria: (1)

randomized controlled trials; (2) the subjects were adultswith hypothyroidism; (3) treatment included levothyrox‐ine or levothyroxine combined with other drugs in thetreatment group; (4) studies were published in Englishor Chinese.

Studies were excluded for meeting the following crite‐ria: (1) duplicate articles or results; (2) obvious dataerrors; (3) case reports, case-control studies, theoreticalresearch, conference reports, systematic reviews, meta-analyses, and other forms of research or comment notdesigned in a randomized controlled manner; (4) irrele‐vant outcomes.

Two investigators independently determined whetherstudies met the inclusion criteria, with a third resolvingany disputes as needed.

Data extraction and quality assessmentFor each eligible study, two categories of information

were extracted: basic information and primary study out‐comes. Basic information relevant to this meta-analysisincluded: author names, year of publication, Jadad score,the detail of disease, test method, interventions, samplesize, age and gender. Primary outcomes relevant to thisanalysis included thyroid stimulating hormone (TSH),systolic blood pressure (SBP), diastolic blood pressure(DBP), total cholesterol (TC), triglycerides (TG), highdensity lipoprotein (HDL), low density lipoprotein(LDL), body mass index (BMI), triiodothyronine (T3),free triiodothyronine (FT3), and free thyroxine (FT4).Study quality was determined on the basis of Jadadscores, which were determined based on how well eachstudy satisfied the following criteria: studies included aspecific statement regarding randomization; the methodused to randomize patients was appropriate; the studywas conducted in a double-blinded manner; the approachto double-blinding was described appropriately; informa‐

tion on any patients that withdrew from or dropped outof the study was provided. A Jadad score <3 indicateda study of low-quality, and thus was associated with asubstantial risk of bias. Data was extracted independentlyby two investigators, with a third resolving any disputesas needed.

Statistical analysisSTATA v10.0 (TX, USA) was used for all analyses.

Heterogeneity in study results was assessed using chi-squared and I2 tests and appropriate analysis models(fixed-effects or random-effects) were determined. Achi-squared p ≤ 0.05 and an I2 > 50% indicated highheterogeneity and the random-effects model was used.A chi-squared p > 0.05 and an I2 ≤ 50% indicated accept‐able heterogeneity and the fixed-effects model wasinstead used. Continuous variables were given as mean ±standard deviations and were compared on the basis ofmean difference (MD), while categorical data were givenas percentages and compared based on relative risk (RR)/odds ratios (ORs). All the indexes were analyzed by MD.

Results

Overview of the Included StudiesWe reviewed a total of 1,106 articles that were identi‐

fied by our initial keyword search, of which 1,015 wereexcluded following title/abstract review. The full text ofthe remaining 91 articles was assessed, excluding 63additional articles that did not meet the study inclusioncriteria. Reasons for exclusion of these studies were:non-randomized studies (31), lack of clinical outcomesof interest (27), duplicate studies (5) and no qualifiedpatients (3). We ultimately identified a total of 25 studies[7-34] that met the inclusion criteria for this meta-analysis, including 1,735 patients. The study selectionprocess is outlined in Fig. 1.

Table 1 summarizes the basic information of eachstudy, including author names, year of publication, detailof disease, test method (TSH, FT4, FT3 and T3), inter‐ventions, sample, age and the Jadad score. The mainJodad score was 4.04, which indicated that the includedstudies were of high quality.

For the kinds of disease, we divided the included stud‐ies into hypothyroidism and subclinical hypothyroidism,then divided into subgroup analysis by intervention:levothyroxine plus liothyronine vs. levothyroxine (L-T4plus L-T3 vs. L-T4), L-T4 vs. placebo, L-T4 plus oraliron vs. oral iron.

TSHTwenty-three studies with 1,627 patients reported

results on TSH.

720 Chen et al.

In patients with hypothyroidism, the value of TSH wassignificantly decreased in the L-T4 plus L-T3 than L-T4(WMD: 0.38, 95%CI: 0.13–0.62), L-T4 plus oral ironthan oral iron (WMD: 4.40, 95%CI: 2.79–6.01), L-T4than placebo (WMD: 3.30, 95%CI: 1.18–5.42). Inpatients with subclinical hypothyroidism, L-T4 signifi‐cantly reduced TSH than placebo (WMD: 4.99, 95%CI:4.29–5.69).

The results are presented in Fig. 2A and Fig. 3A.

SBP and DBPEight studies with 824 patients reported results

on SBP and DBP. In the patients with hypothyroid‐ism, there was no significant difference in SBP(WMD:–0.419, 95%CI: –6.371–5.534) and DBP (WMD:2.075, 95%CI: –2.375–6.524) between the L-T4 plusL-T3 group and the L-T4 group. In patients with subclin‐ical hypothyroidism, there was no significant differencein DBP (WMD: 0.867, 95%CI: –1.073–2.808) betweenthe L-T4 group and the placebo group. Compared withplacebo, L-T4 significantly decreased SBP (WMD:2.540, 95%CI: 1.332–3.748).

The results are presented in Table 2 and Table 3.

T3/FT3 and FT4Seven studies with 376 patients reported results on T3.

In patients with hypothyroidism, there was no significantdifference in T3 between the L-T4 plus L-T3 group andthe L-T4 group (WMD: 1.75, 95%CI: –0.13–3.63), andbetween the L-T4 group and the placebo group (WMD:–80.64, 95%CI: –175.86–14.58). In patients with sub‐clinical hypothyroidism, the value of T3 was signifi‐cantly decreased in the L-T4 group vs. the placebo group(WMD: 38.500, 95%CI: 23.408–53.592).

Eleven studies with 1,019 patients reported results onFT3. In patients with hypothyroidism, the value of FT3was significantly increased in the L-T4 plus L-T3 groupvs. the L-T4 group (WMD: –1.700, 95%CI: –2.337––1.063). In patients with subclinical hypothyroidism, thevalue of FT3 was significantly increased in the L-4group vs. the placebo group (WMD: –0.46, 95%CI:–0.65– –0.27).

Twenty-two studies with 1,601 patients reportedresults on FT4. In patients with hypothyroidism, therewas no significant difference in FT4 (WMD: 0.71,95%CI: –0.06–1.48) between the L-T4 plus L-T3group and the L-T4 group; compared with placebo, L-T4significantly increased FT4 (WMD: –19.30, 95%CI:–34.08– –4.53). In patients with subclinical hypo‐thyroidism, compared with placebo, L-T4 significantlyincreased FT4 (WMD: –2.49, 95%CI: –3.42– –1.56).Compared with oral iron, L-T4 plus oral iron signifi‐cantly decreased FT4 (WMD: 1.59, 95%CI: 0.40–2.78).

The results are presented in Fig. 2B–C, Fig. 3B–C,Table 2 and Table 3.

TC and TGFourteen studies with 828 patients reported results

on TC. Twelve studies with 764 patients reported resultson TG. In patients with hypothyroidism, there was nosignificant difference in TC (WMD: 20.142, 95%CI:–21.426–61.709) and TG (WMD: –3.078, 95%CI:–26.499–20.344) between the L-T4 plus L-T3 group andthe L-T4 group. In patients with subclinical hypothyroid‐ism, there was no significant difference in TG (WMD:–3.165, 95%CI: –8.397–2.067) between the L-T4 groupand the placebo group. Compared with placebo, L-T4significantly decreased TC (WMD: 5.688, 95%CI:

Fig. 1 Literature search and selection strategy.

Levothyroxine for hypothyroidism 721

Tabl

e 1

The

basic

cha

ract

erist

ics d

escr

iptio

n of

incl

uded

stud

ies

Stud

yJa

dad

scor

eD

iseas

eTe

st M

etho

d of

TSH

, FT4

, FT3

and

T3

Inte

rven

tion

No.

of

patie

nts

Age

Gen

der

TC

TC

TC

TC

Robe

rtK

rysia

k 20

182

youn

g w

omen

with

hypo

thyr

oidi

sm

chem

ilum

ines

cenc

e us

ing

acrid

iniu

mes

ter t

echn

olog

y (A

DV

IA C

enta

ur X

PIm

mun

oass

ay S

yste

m, S

iem

ens

Hea

lthca

re D

iagn

ostic

s, M

unic

h,G

erm

any)

L-T4

plu

s L-T

3: 7

5–10

0 μg

/dai

lyle

voth

yrox

ine

was

repl

aced

by

a pr

epar

atio

nco

ntai

ning

bot

hle

voth

yrox

ine

and

lioth

yron

ine

in a

wei

ght

prop

ortio

n of

5:1

L-T4

75–

100

μg/

daily

17

20

3130

17

F20

F

Yong

Che

n20

184

early

‑sta

ge d

iabe

ticne

phro

path

y pa

tient

s with

subc

linic

alhy

poth

yroi

dism

Plas

ma

FT3,

FT4

, and

TSH

wer

ede

term

ined

by

radi

oim

mun

oass

ay(n

orm

al ra

nge:

FT3

, 1.7

1–3.

71 p

g/m

L;FT

4, 0

.70

FT3,

1.7

1L; T

SH, 0

.3–4

.94

μIU

/mL)

(Lin

co R

esea

rch,

St.

Char

les,

MO

, USA

).

L-T4

12.

5 μg

/day

and

grad

ually

incr

ease

dac

cord

ing

to c

linic

alsy

mpt

oms a

nd la

bora

tory

resu

lts.

plac

ebo

4341

5156

15M

14M

Julia

naK

amin

ski

2016

5A

dults

with

prim

ary

hypo

thyr

oidi

sm—

LT4/

LT3

ther

apy

(75/

15 μ

g/da

y)

L-T4

125

or 1

50μg

/day

1517

46.1

39.5

15F

15F

Men

g Zh

ao20

165

Mild

Sub

clin

ical

Hyp

othy

roid

ismra

dioi

mm

unoa

ssay

(Lin

co R

esea

rch,

St.

Char

les,

MO

, USA

)L-

T4 1

2.5–

125

μg o

nce

daily

plac

ebo

210

159

54.9

855

.44

57M

42M

Peng

Liu

201

55

patie

nts w

ith e

arly

type

2di

abet

ic n

ephr

opat

hy a

ndsu

bclin

ical

hypo

thyr

oidi

sm

TSH

by

IRM

A, T

4 by

RIA

L-T4

, sta

rted

from

a lo

wdo

se o

f 12.

5 μg

/d, a

ndgr

adua

lly in

crea

sed

acco

rdin

g to

clin

ical

sym

ptom

s and

labo

rato

ryte

st re

sults

plac

ebo

6059

4949

13M

14M

Laily

Naj

afi

2015

5pa

tient

s with

subc

linic

alhy

poth

yroi

dism

AD

VIA

Cen

taur

® im

mun

oass

ay sy

stem

,Si

emen

s Hea

lthca

re D

iagn

ostic

sL-

T4 1

00 u

g da

ilypl

aceb

o30

3032

.47

36.0

725

F26

F

Chan

tal J

. N.

Verk

leij

2013

a3

hypo

thyr

oidi

sm a

ndhy

perth

yrox

inae

mia

AD

VIA

Cen

taur

® im

mun

oass

ay sy

stem

,Si

emen

s Hea

lthca

re D

iagn

ostic

sL-

T4 0

.3 m

g pe

r day

plac

ebo

1620

3052

9M7M

Chan

tal J

. N.

Verk

leij

2013

b3

hypo

thyr

oidi

sm a

ndhy

perth

yrox

inae

mia

imm

unoc

hem

ilum

ines

cenc

e (Im

mul

ite20

00; D

PC, D

iagn

ostic

Pro

duct

sCo

rpor

atio

n, U

SA).

L-T4

0.4

5 m

g pe

r day

plac

ebo

1220

2952

6M7M

Mon

ica

Dia

sCa

bral

201

12

fem

ale

patie

nts w

ith m

ildsu

bclin

ical

hypo

thyr

oidi

sm

imm

unoc

hem

ilum

ines

cenc

e m

etho

d(“

Imm

ulite

”, D

iagn

ostic

Pro

duct

sCo

rpor

atio

n, L

os A

ngel

es, U

SA)

L-T4

0.7

5 m

g/kg

/day

plac

ebo

1418

43.3

647

.59

14F

18F

722 Chen et al.

Tabl

e 1

Cont

.

Stud

yJa

dad

scor

eD

iseas

eTe

st M

etho

d of

TSH

, FT4

, FT3

and

T3

Inte

rven

tion

No.

of

patie

nts

Age

Gen

der

TC

TC

TC

TC

Vale

ntin

e V.

Fade

yev

2010

3pr

imar

y hy

poth

yroi

dism

Com

mer

cial

ly a

vaila

ble

high

-cap

acity

,ra

ndom

acc

ess i

mm

unoa

ssay

kits

wer

eus

ed to

mea

sure

FT4

, FT3

, and

TSH

leve

ls (C

hiro

n D

iagn

ostic

s Co.

, Wal

pole

,M

A, U

SA)

25 μ

g L-

T4 a

nd 1

2.5

μg o

fL-

T3L-

T41.

6 μg

/kg

16

20

4043

—

—

Tosh

iki

Nag

asak

i 200

94

fem

ale

subc

linic

alhy

poth

yroi

d

Seru

m F

T3 a

nd F

T4 c

once

ntra

tions

wer

em

easu

red

by sp

ecifi

c RI

A (T

echn

o-G

enet

ics R

ecor

dati,

Mila

n, It

aly)

. Ser

umTS

H le

vels

wer

e de

term

ined

by

anul

trase

nsiti

ve im

mun

orad

iom

etric

ass

ay(IR

MA

) met

hod

(Cis

Dia

gnos

tici,

Tron

zano

Ver

celle

se, I

taly

).

L-T4

12.

5 μg

–37.

5 μg

plac

ebo

4847

64.4

66

——

Hak

anCi

nem

re 2

008

5Iro

n-D

efic

ient

Sub

clin

ical

Hyp

othy

roid

Pat

ient

s

Seru

m T

SH c

once

ntra

tion

(refe

renc

era

nge

0.1

to 4

.0 m

IU/L

) was

mea

sure

dby

a se

nsiti

ve im

mun

omet

ric a

ssay

(Del

fia; W

alla

c, T

urku

, Fin

land

). Fr

ee T

4(8

.0 to

23.

0 pm

ol/L

) and

tota

l T3

(1.2

to3.

1 nm

ol/L

) wer

e de

term

ined

by

mic

ropa

rticl

e en

zym

e im

mun

oass

ays

IMx

(Abb

ott,

Baar

, Sw

itzer

land

).

L-T4

75

μg/d

plu

s ora

liro

n 24

0 m

g/d

oral

iron

240

mg/

d25

2535

3822

M22

M

Salm

an R

azvi

2007

5Su

bclin

ical

Hyp

othy

roid

ismen

zym

e im

mun

oass

ay k

its (A

bbot

tLa

bora

torie

s, A

bbot

t Par

k, rl

L-T4

100

mu

plac

ebo

5050

53.5

54.2

10M

8M

F. M

ON

ZAN

I20

044

Subc

linic

alH

ypot

hyro

idism

TSH

con

cent

ratio

ns (u

sing

a th

irdge

nera

tion

TSH

ass

ay, A

xsym

II, A

bbot

tLa

bora

torie

s, Ch

icag

o, IL

), an

d fre

e T4

,fre

e T3

(usin

g flu

ores

cent

imm

unoa

ssay

s, A

xsym

, Abb

otLa

bora

torie

s)

L-T4

70

says

, Ax

plac

ebo

2322

——

——

Chris

tian

Mei

er 2

004

5su

bclin

ical

hypo

thyr

oidi

sm

Seru

m T

SH c

once

ntra

tion

(refe

renc

era

nge

0.3/

4.0

mU

/L) w

as m

easu

red

by a

nim

mun

omet

ric a

ssay

(Del

fia, W

alla

c,Tu

rku,

Fin

land

). Fr

ee T

4 (re

fere

nce

rang

e 8.

0/23

.0 p

mol

/L) a

nd to

tal T

3(re

fere

nce

rang

e 1.

2/3.

1 nm

ol/L

) wer

ede

term

ined

by

mic

ropa

rticl

e en

zym

eim

mun

oass

ays I

Mx

(Abb

ott,

Dia

gnos

ticD

ivisi

on, C

hica

go, I

L). t

Hcy

(ref

eren

cera

nge

5/15

mm

ol/L

) was

mea

sure

d w

ithth

e IM

x fu

lly a

utom

ated

fluo

resc

ence

pola

rizat

ion

imm

unoa

ssay

(Abb

ott).

L-T4

85.

5 flu

ores

cenc

epl

aceb

o33

3357

.157

.1—

—


Tabl

e 1

Cont

.

Stud

yJa

dad

scor

eD

iseas

eTe

st M

etho

d of

TSH

, FT4

, FT3

and

T3

Inte

rven

tion

No.

of

patie

nts

Age

Gen

der

TC

TC

TC

TC

Patri

ck W

.Cl

yde

2003

5pr

imar

y hy

poth

yroi

dism

Seru

m F

T3 a

nd F

T4 c

once

ntra

tions

wer

em

easu

red

by sp

ecifi

c RI

A (T

echn

o-G

enet

ics R

ecor

dati,

Mila

n, It

aly)

. Ser

umTS

H le

vels

wer

e de

term

ined

by

anul

trase

nsiti

ve im

mun

orad

iom

etric

ass

ay(IR

MA

) met

hod

(Cis

Dia

gnos

tici,

Tron

zano

Ver

celle

se, I

taly

)

L-T4

50

(Cis

Dia

gnos

tici,

twic

e da

ily

L-T4

25

usL-

T3ag

nosti

c

2222

43

.145

.2

19F

17F

A. M

.SA

WK

A 2

003

4Pa

tient

s with

Hyp

othy

roid

ism

The

seru

m T

SH c

once

ntra

tion

(refe

renc

era

nge,

0.1

cen

tratio

n (re

fere

nce

rang

e, n

oVe

mm

unom

etric

ass

ay (D

elfia

, Wal

lac,

Inc.

, Tur

ku, F

inla

nd).

Free

T4

(8.0

em

mul

lese

, Ita

ly) o

tt).,

rland

) (1.

2,Tu

rku,

Fin

land

) wer

e de

term

ined

by

mic

ropa

rticl

e en

zym

e im

mun

oass

ays

(IMx,

Abb

ott L

abor

ator

ies,

Inc.

,Ch

icag

o, IL

).

L-T4

25

bott

Labo

rato

ries

L-T4

50

us

2020

4549

.5

3M1M

Mirj

am C

hrist

-Cr

ain

2003

4hy

poth

yroi

dism

Seru

m F

T3 a

nd F

T4 le

vels

wer

em

easu

red

by sp

ecifi

c RI

A (T

echn

o-G

enet

ics R

ecor

dati,

Mila

n, It

aly)

. TSH

was

det

erm

ined

with

an

ultra

sens

itive

imm

unor

adio

met

ric a

ssay

met

hod

(Cis

Dia

gnos

tici,

Tron

zano

Ver

celle

se, I

taly

).

L-T4

plac

ebo

3132

——

——

NA

DIA

CARA

CCIO

2002

4Su

bclin

ical

Hyp

othy

roid

ism

TSH

and

FT4

wer

e m

easu

red

byim

mun

orad

iom

etric

ass

ay u

sing

a TS

H-

CTK

-3 k

it (D

iaSo

rin S

alug

gia,

Ital

y) a

ndan

FT4

-CTK

kit

(Dia

Sorin

).

L-T4

67.

5T4

wpl

aceb

o24

25—

——

—

CHRI

STIA

NM

EIER

200

14

Subc

linic

alH

ypot

hyro

idism

third

-gen

erat

ion

elec

troch

emilu

min

esce

nce

imm

unoa

ssay

(Gm

bh, D

-682

98, r

oche

dia

gnos

tics,

man

nhei

m G

erm

any)

L-T4

85.

5 er

atio

nel

ectro

chem

ilum

ines

cepl

aceb

o33

3357

.157

.1—

—

FABI

OM

ON

ZAN

I20

014

Subc

linic

alH

ypot

hyro

idism

TSH

by

imm

unoc

hem

ilum

inom

etric

assa

y (L

IAIS

ON

TSH

, BY

K G

ULD

ENIT

ALI

A, M

ILA

N, I

TALY

), FT

3 A

ND

FT4

BY R

AD

IOIM

MU

NO

ASS

AY(L

IAIS

ON

T4

AN

D T

3, B

YK

GU

LDEN

ITA

LIA

)

L-T4

0.0

5 m

g da

ilypl

aceb

o10

1034

.329

.29F

9F

724 Chen et al.

Tabl

e 1

Cont

.

Stud

yJa

dad

scor

eD

iseas

eTe

st M

etho

d of

TSH

, FT4

, FT3

and

T3

Inte

rven

tion

No.

of

patie

nts

Age

Gen

der

TC

TC

TC

TC

Chun

g-H

oon

Kim

201

03

subc

linic

alhy

poth

yroi

dism

imm

unoc

hem

ilum

ines

cent

pro

cedu

res

(DPC

: dia

gnos

tic p

rodu

cts c

orpo

ratio

n/im

mul

ite 2

000)

L-T4

50

hem

ipl

aceb

o

3232

36

36.1

—

—

Ash

raf H

any

Abd

el R

ahm

an20

092

subc

linic

alhy

poth

yroi

dism

imm

unoc

hem

ilum

ines

cent

ass

ays

(imm

ulite

200

0, d

iagn

ostic

pro

duct

scr

op.,

Los A

ngel

es, C

alifo

rnia

)L-

T4 5

0–10

0 m

cg d

aily

plac

ebo

3535

31.2

30

——

Mai

jdVa

lizad

eh20

095

prim

ary

hypo

thyr

oidi

sm

chem

ilum

ines

cenc

e us

ing

acrid

iniu

mes

ter t

echn

olog

y (A

DV

IA C

enta

ur X

PIm

mun

oass

ay S

yste

m, S

iem

ens

Hea

lthca

re D

iagn

ostic

s, M

unic

h,G

erm

any)

L-T4

50

umin

esce

nce

usin

g ac

ridin

iuL-

T4 1

00 s

3030

39.2

38.8

24F

24F

P. F

. S.

Teix

eira

200

85

Subc

linic

alH

ypot

hyro

idism

Plas

ma

FT3,

FT4

, and

TSH

wer

ede

term

ined

by

radi

oim

mun

oass

ay(n

orm

al ra

nge:

FT3

, 1.7

1–3.

71 p

g/m

L;FT

4, 0

.70

FT3,

1.7

1 L;

TSH

, 0.3

–4.

94 μ

IU/m

L) (L

inco

Res

earc

h, S

t.Ch

arle

s, M

O, U

SA).

L-T4

25–

75 S

tpl

aceb

o18

2052

.546

.6—

—

Héc

tor F

.Es

coba

r-M

orre

ale

2005

5Pr

imar

y H

ypot

hyro

idism

TSH

was

det

erm

ined

in d

uplic

ates

by

Imm

ulite

® 2

000

chem

ilum

ines

cenc

eTS

H th

ird-g

ener

atio

n ki

t (D

iagn

ostic

Prod

ucts

Corp

orat

ion,

CA

, USA

;re

fere

nce

valu

es [R

V] 0

.4–4

.0 m

UI/L

;se

nsiti

vity

0.0

02 m

UI/L

; int

ra-a

ssay

coef

ficie

nt o

f var

iatio

n [C

V],

3.8–

12.5

%;

and

inte

r-ass

ay C

V, 4

.6–1

2.5%

). Fr

ee T

4w

as m

easu

red

in d

uplic

ates

by

Imm

ulite

®

2000

che

milu

min

esce

nce

enzy

mel

inke

dim

mun

osor

bent

ass

ay k

it (D

iagn

ostic

Prod

ucts

Corp

orat

ion,

CA

, USA

; RV,

0.8–

1.9

ng/d

L; se

nsiti

vity

, 0.1

5 ng

/dL;

intra

-ass

ay C

V, 4

.4%

–7.5

%; a

nd in

ter-

assa

y CV

, 4.8

%–9

.0%

). To

tal T

3 w

asm

easu

red

in d

uplic

ates

by

Imm

ulite

®

2000

che

milu

min

esce

nce

enzy

mel

inke

dim

mun

osor

bent

ass

ay k

it (D

iagn

ostic

Prod

ucts

Corp

orat

ion,

CA

, USA

; RV,

82–1

79 n

g/dL

; sen

sitiv

ity, 1

9 ng

/dL;

intra

-ass

ay C

V, 4

.4%

–12%

; and

inte

r-as

say

CV, 5

.3%

–15%

).

L-T4

75

μg p

lus L

-T3

5 μg

L-T4

100

μg14

14—

——

—


Fig. 2 A, Forest plot for TSH; B, Forest plot for FT4; C, Forest plot for T3; in patients with hypothyroidism.

726 Chen et al.

Fig. 3 A, Forest plot for TSH; B, Forest plot for FT4; C, Forest plot for FT3; in patients with subclinical hypothyroidism.


0.433–10.944).The results are presented in Table 2 and Table 3.

HDL and LDLTwelve studies with 764 patients reported results on

HDL and LDL. In patients with hypothyroidism, therewas no significant difference in HDL (WMD: 1.003,95%CI: –1.643–3.648) and LDL (WMD: 16.090, 95%CI:–14.189–46.369) between the L-T4 plus L-T3 group andthe L-T4 group. In patients with subclinical hypothyroid‐ism, there was no significant difference in HDL (WMD:–0.017, 95%CI: –0.076–0.042) and LDL (WMD: 5.077,95%CI: –0.769–10.922) between the L-T4 group and theplacebo group.


BMISeven studies with 397 patients reported results on

BMI. In patients with hypothyroidism, there was no

significant difference in BMI (WMD: 2.109, 95%CI:–0.030–4.248) between the L-T4 plus L-T3 group andthe L-T4 group. In the patients with subclinical hypo‐thyroidism, there was no significant difference in BMI(WMD: 0.091, 95%CI: –0.087–0.268) between the L-T4group vs. the placebo group.


Quality and bias assessmentAn assessment of study quality and risk of bias was

performed using multiple complementary methodsincluding funnel plots, Begg’s and Mazumdar’s ranktest, and Egger’s test. There was clear symmetry in thelog WMD funnel plot for TSH in studies on patients withhypothyroidism, suggesting a low publication bias risk(Fig. 4). The results of Begg’s and Mazumdar’s rank test(Z = 1.61, p = 0.107) and Egger’s test (p = 0.781) allsuggested that there was not any significant risk of biasamong the study results. There was clear symmetry in

Table 2 The others indexes results of patients with hypothyroidism

Index N (case/control) Interventions WMD (95%CI) p* I2 p#p value

Begg’s Egger’s

SBP 37/39 L-T4 plus L-T3 vs. L-T4 –0.419 (–6.371, 5.534) 0.676 0.0% 0.890 1.000 —

DBP 37/39 L-T4 plus L-T3 vs. L-T4 2.075 (–2.375, 6.524) 0.247 25.3% 0.361 1.000 —

FT3 34/34 L-T4 plus L-T3 vs. L-T4 –1.700 (–2.337, –1.063) 0.344 0.0% 0.000 1.000 —

TC 97/103 L-T4 plus L-T3 vs. L-T4 20.142 (–21.426, 61.709) 0.000 96.8% 0.342 0.806 0.021

TG 67/69 L-T4 plus L-T3 vs. L-T4 –3.078 (–26.499, 20.344) 0.825 0.0% 0.797 1.000 0.612

HDL 67/69 L-T4 plus L-T3 vs. L-T4 1.003 (–1.643, 3.648) 0.963 0.0% 0.457 1.000 0.903

LDL 82/86 L-T4 plus L-T3 vs. L-T4 16.090 (–14.189, 46.369) 0.000 94.7% 0.298 1.000 0.163

BMI 32/37 L-T4 plus L-T3 vs. L-T4 2.109 (–0.030, 4.248) 0.769 0.0% 0.053 1.000 —

* p value of Heterogeneity chi-squared# p value of Pooled statistic

Table 3 The others indexes results of patients with subclinical hypothyroidism

Index N (case/control) Interventions WMD (95%CI) p* I2 p#p value

Begg’s Egger’s

SBP 401/347 L-T4 vs. placebo 2.540 (1.332, 3.748) 0.199 31.5% 0.000 1.000 0.866

DBP 401/347 L-T4 vs. placebo 0.867 (–1.073, 2.808) 0.008 67.9% 0.381 0.060 0.010

T3 66/66 L-T4 vs. placebo 38.500 (23.408, 53.592) 0.004 87.9% 0.000 1.000 —

TC 313/315 L-T4 vs. placebo 5.688 (0.433, 10.944) 0.000 79.5% 0.034 0.917 0.077

TG 313/315 L-T4 vs. placebo –3.165 (–8.397, 2.067) 0.000 74.5% 0.236 0.602 0.348

HDL 313/315 L-T4 vs. placebo –0.017 (–0.076, 0.042) 0.423 1.3% 0.575 0.251 0.010

LDL 313/315 L-T4 vs. placebo 5.077 (–0.769, 10.922) 0.000 87.1% 0.089 0.602 0.163

BMI 165/163 L-T4 vs. placebo 0.091 (–0.087, 0.268) 0.945 0.0% 0.317 0.086 0.207

* p value of Heterogeneity chi-squared# p value of Pooled statistic

728 Chen et al.

the log WMD funnel plot for FT4 in studies on patientswith subclinical hypothyroidism, suggesting a low publi‐cation bias risk (Fig. 5). The results of Begg’s andMazumdar’s rank test (Z = 1.39, p = 0.166) and Egger’stest (p = 0.164) all suggested that there was not any sig‐nificant risk of bias among the study results.

Discussion

Hypothyroidism is a systemic hypometabolic syn‐drome caused by a decrease in thyroid hormone synthe‐sis and secretion or a lack of biological effects. Theincidence of the disease is 3%–7%, mostly in women,and is positively correlated with age. Hypothyroidism isa systemic metabolic syndrome caused by a decrease inthyroid hormone synthesis and secretion or a decrease intissue function, which is mainly divided into clinical andsubclinical hypothyroidism. The prevalence of hypothy‐roidism is related to TSH diagnostic cut-off point value,

Fig. 4 Funnel plot analysis of the included studies of patientswith hypothyroidism.

Fig. 5 Funnel plot analysis of the included studies of patientswith subclinical hypothyroidism.

age, gender, race and other factors, among which theprevalence of subclinical hypothyroidism is higher thanthat of clinical hypothyroidism. TSH, FT4 and TT4 arethe first line parameters for diagnosis of clinical hypo‐thyroidism. The diagnostic criteria for subclinical hypo‐thyroidism are usually TSH levels higher than thereference range combined with normal FT4 levels.Hyperlipidemia is a common complication of hypothy‐roidism and can increase the risk of cardiovascular andcerebrovascular diseases. Therefore, it is of great signifi‐cance to improve the long-term prognosis of patientswith hypothyroidism. The relationship between hypo‐thyroidism and abnormal lipid metabolism is not clear,and many studies suggest that the mechanism is mainlyrelated to TSH elevation and decline in FT3 and FT4.Basic studies have confirmed that TSH is involved inlipid metabolism, and TSH elevations in hypothyroidismare closely related to abnormal lipid metabolism. Thy‐roid hormone is involved in regulating blood lipid syn‐thesis, promote cholesterol metabolism and mobilizationprocess and its metabolite excretion. Hypothyroidismcauses elevations in TC and TG. The number and activ‐ity of LDL receptors in hepatocytes are positively corre‐lated with thyroid function. Hypothyroidism can lead toimpaired receptor-dependent LDL lowering pathways,causing LDL levels to rise. Decreased lipase activity inhypothyroidism may be the direct cause of elevated TC,TG and LDL levels.

L-T4 is the main alternative drug for the treatment ofhypothyroidism. Long-term experience has proved thatL-T4 has the advantages of reliable efficacy, few adversereactions, good compliance, good intestinal absorption,long serum half-life and low treatment cost. Patientswith hypothyroidism are deficient in endogenous thyroidhormones. The normal thyroid gland secretes about 85 gT4 every day. About 80% of T3 (about 26 g) is convertedfrom peripheral T4, and only 20% (about 6.5 g) isdirectly secreted by the thyroid. Currently, it is generallybelieved that although T4 is the main hormone secretedby the thyroid, the thyroid hormone mainly binds to itsnuclear receptor by T3 in the peripheral tissues. Therationale for the treatment of hypothyroidism with L-T4is the conversion of exogenous thyroxine (T4) fromperipheral tissues to the active metabolite T3. The gas‐trointestinal absorption rate of L-T4 tablets can reach70%–80%. L-T4 tablets have a half-life of about 7 days,and stable serum T4 and T3 levels can be obtained byadministration once daily. The dose of L-T4 needs to beindividualized, depending on the patient’s condition, age,and weight. L-T4 treatment must undergo transformationfrom T4 to T3, and the theoretical advantage of L-T3treatment is that it can avoid this process and directlyenable active hormones to play their role. However, the


drawback of L-T3 monotherapy is the absence of thesubstrate T4, and the level of T3 in circulation and tissueis completely dependent on exogenous hormone replace‐ment therapy. The dose and duration of L-T3 should bestrictly adhered to. Over-dose or under-dose of L-T3 canincrease the risk of heart and bone side effects.

In a previous similar meta-analysis, Li X et al. [35]included 12 trials and found that levothyroxine treatmenthas clear benefits on TC and LDL-C in subclinical hypo‐thyroidism patients, including those with mild subclini‐cal hypothyroidism. Yamamoto et al. [36] concluded thatthere was no evidence of beneficial effects of levothyr‐oxine therapy on obstetrical, neonatal, childhood IQ orneurodevelopmental outcomes. Current trial evidencedoes not support the treatment of subclinical hypothyr‐oidism diagnosed in pregnancy.

In our study, in patients with hypothyroidism: com‐pared with L-T4, L-T4 plus L-T3 would significantlydecrease the value of TSH (WMD: 0.38, 95%CI: 0.13–0.62) and increased the value of FT3 (WMD: –1.700,95%CI: –2.337– –1.063). There was no significant dif‐ference in SBP (WMD: –0.419, 95%CI: –6.371–5.534),DBP (WMD: 2.075, 95%CI: –2.375–6.524), T3 (WMD:1.75, 95%CI: –0.13–3.63), FT4 (WMD: 0.71, 95%CI:–0.06–1.48), TC (WMD: 20.142, 95%CI: –21.426–61.709), TG (WMD: –3.078, 95%CI: –26.499–20.344),HDL (WMD: 1.003, 95%CI: –1.643–3.648), LDL(WMD: 16.090, 95%CI: –14.189–46.369) and BMI(WMD: 2.109, 95%CI: –0.030–4.248) between the L-T4plus L-T3 group and the L-T4 group. The value of TSHwas significantly decreased in the L-T4 plus oral irongroup versus the oral iron group (WMD: 4.40, 95%CI:2.79–6.01). There was no significant difference in T3 inthe L-T4 group versus the placebo group (WMD: –80.64,95%CI: –175.86–14.58). Compared with placebo, L-T4significantly increased FT4 (WMD: –19.30, 95%CI:–34.08– –4.53) and decreased TSH (WMD: 3.30,

95%CI: 1.18–5.42).In patients with subclinical hypothyroidism: compared

with placebo, L-T4 significantly decreased SBP (WMD:2.540, 95%CI: 1.332–3.748), TSH (WMD: 4.99, 95%CI:4.29–5.69), T3 (WMD: 38.500, 95%CI: 23.408–53.592)and TC (WMD: 5.688, 95%CI: 0.433–10.944) andincreased FT3 (WMD: –0.46, 95%CI: –0.65– –0.27),FT4 (WMD: –2.49, 95%CI: –3.42– –1.56). There was nosignificant difference in DBP (WMD: 0.867, 95%CI:–1.073–2.808), TG (WMD: –3.165, 95%CI: –8.397–2.067), BMI (WMD: 0.091, 95%CI: –0.087–0.268),HDL (WMD: –0.017, 95%CI: –0.076–0.042) and LDL(WMD: 5.077, 95%CI: –0.769–10.922) between the L-T4 group and the placebo group. Compared with oraliron, L-T4 plus oral iron significantly decreased FT4(WMD: 1.59, 95%CI: 0.40–2.78).

However, there are certain limitations to the presentanalysis, which are as follows: (1) the number of studiesincluded in the subgroup analysis is limited. (2) individ‐ual studies were varied in exclusion/inclusion criteria. (3)we only included RCT. (4) the severity of hypothyroid‐ism varied among the studies. (5) the kits used to meas‐ure TSH, FT4 and FT3 were different among the studies.Kits by different manufacturers had significant effects onthe results of TSH, FT4 and FT3. For the limited numberof included studies, we have already carried out sub‐group analysis by diseases and interventions. If we con‐tinued subgroup analysis by kits, there would be fewerstudies in the subgroup, so the results would have nomore clinical significance. (6) The doses of L-T4 and L-T3 were varied among the studies. In most included stud‐ies, the doses of L-T4 and L-T3 started from a low dose,and gradually increased according to clinical symptomsand laboratory test results, so it’s not a fixed dose thatmay have different effects on the thyroid hormones. (7)Pooled data were analyzed, as individual patient datawas not available, precluding more in-depth analyses.

References

1. Wassner AJ (2017) Pediatric hypothyroidism: diagnosisand treatment. Paediatr Drugs 19: 291–301.

2. Shan Z, Teng W (2019) Thyroid hormone therapy ofhypothyroidism in pregnancy. Endocrine 66: 35–42.

3. Magri F, Chiovato L, Croce L, Rotondi M (2019) Thyroidhormone therapy for subclinical hypothyroidism. Endo‐crine 66: 27–34.

4. Duntas LH, Yen PM (2019) Diagnosis and treatment ofhypothyroidism in the elderly. Endocrine 66: 63–69.

5. Bauer AJ, Wassner AJ (2019) Thyroid hormone therapy incongenital hypothyroidism and pediatric hypothyroidism.Endocrine 66: 51–62.

6. Fletcher B (2019) Junior doctors are embarrassing them‐

selves by opposing medical associate professionals. BMJ365: l4255.

7. Krysiak R, Szkrobka W, Okopien B (2018) Sexual func‐tion and depressive symptoms in young women withhypothyroidism receiving levothyroxine/liothyroninecombination therapy: a pilot study. Curr Med Res Opin34: 1579–1586.

8. Chen Y, Wu G, Xu M (2018) The effect of L-thyroxinesubstitution on oxidative stress in early-stage diabeticnephropathy patients with subclinical hypothyroidism: arandomized double-blind and placebo-controlled study. IntUrol Nephrol 50: 97–103.

9. Nazarpour S, Ramezani Tehrani F, Simbar M, Tohidi M,

730 Chen et al.

Minooee S, et al. (2018) Effects of levothyroxine on preg‐nant women with subclinical hypothyroidism, negative forthyroid peroxidase antibodies. J Clin Endocrinol Metab103: 926–935.

10. Kaminski J, Miasaki FY, Paz-Filho G, Graf H, CarvalhoGA (2016) Treatment of hypothyroidism with levothyrox‐ine plus liothyronine: a randomized, double-blind, cross‐over study. Arch Endocrinol Metab 60: 562–572.

11. Zhao M, Liu L, Wang F, Yuan Z, Zhang X, et al. (2016) Aworthy finding: decrease in total cholesterol and low-density lipoprotein cholesterol in treated mild subclinicalhypothyroidism. Thyroid 26: 1019–1029.

12. Liu P, Liu R, Chen X, Chen Y, Wang D, et al. (2015) Canlevothyroxine treatment reduce urinary albumin excretionrate in patients with early type 2 diabetic nephropathy andsubclinical hypothyroidism? A randomized double-blindand placebo-controlled study. Curr Med Res Opin 31:2233–2240.

13. Najafi L, Malek M, Hadian A, Ebrahim Valojerdi A,Khamseh ME, et al. (2015) Depressive symptoms inpatients with subclinical hypothyroidism—the effect oftreatment with levothyroxine: a double-blind randomizedclinical trial. Endocr Res 40: 121–126.

14. Verkleij CJ, Stuijver DJ, van Zaane B, Squizzato A,Brandjes DP, et al. (2013) Thrombin-activatable fibrinoly‐sis inhibitor in hypothyroidism and hyperthyroxinaemia.Thromb Haemost 109: 214–220.

15. Cabral MD, Teixeira P, Soares D, Leite S, Salles E, et al.(2011) Effects of thyroxine replacement on endothelialfunction and carotid artery intima-media thickness infemale patients with mild subclinical hypothyroidism.Clinics (Sao Paulo) 66: 1321–1328.

16. Kim CH, Ahn JW, Kang SP, Kim SH, Chae HD, et al.(2011) Effect of levothyroxine treatment on in vitro fertili‐zation and pregnancy outcome in infertile women withsubclinical hypothyroidism undergoing in vitro fertiliza‐tion/intracytoplasmic sperm injection. Fertil Steril 95:1650–1654.

17. Fadeyev VV, Morgunova TB, Melnichenko GA, Dedov II(2010) Combined therapy with L-thyroxine and L-triiodothyronine compared to L-thyroxine alone in thetreatment of primary hypothyroidism. Hormones 9: 245–252.

18. Abdel Rahman AH, Aly Abbassy H, Abbassy AA (2010)Improved in vitro fertilization outcomes after treatment ofsubclinical hypothyroidism in infertile women. EndocrPract 16: 792–797.

19. Kiyici S, Gul OO, Baskan EB, Hacioglu S, Budak F, et al.(2010) Effect of levothyroxine treatment on clinical symp‐toms and serum cytokine levels in euthyroid patients withchronic idiopathic urticaria and thyroid autoimmunity.Clin Exp Dermatol 35: 603–607.

20. Valizadeh M, Seyyed-Majidi MR, Hajibeigloo H,Momtazi S, Musavinasab N, et al. (2009) Efficacy ofcombined levothyroxine and liothyronine as comparedwith levothyroxine monotherapy in primary hypothyroid‐ism: a randomized controlled trial. Endocr Res 34: 80–89.

21. Mainenti MR, Vigario PS, Teixeira PF, Maia MD,Oliveira FP, et al. (2009) Effect of levothyroxine replace‐ment on exercise performance in subclinical hypothyroid‐ism. J Endocrinol Invest 32: 470–473.

22. Nagasaki T, Inaba M, Yamada S, Shirakawa K, Nagata Y,et al. (2009) Decrease of brachial-ankle pulse wave veloc‐ity in female subclinical hypothyroid patients duringnormalization of thyroid function: a double-blind,placebo-controlled study. European Eur J Endocrinol 160:409–415.

23. Cinemre H, Bilir C, Gokosmanoglu F, Bahcebasi T (2009)Hematologic effects of levothyroxine in iron-deficientsubclinical hypothyroid patients: a randomized, double-blind, controlled study. J Clin Endocrinol Metab 94: 151–156.

24. Teixeira PF, Reuters VS, Ferreira MM, Almeida CP, ReisFA, et al. (2008) Treatment of subclinical hypothyroidismreduces atherogenic lipid levels in a placebo-controlleddouble-blind clinical trial. Horm Metab Res 40: 50–55.

25. Razvi S, Ingoe L, Keeka G, Oates C, McMillan C, et al.(2007) The beneficial effect of L-thyroxine on cardiovas‐cular risk factors, endothelial function, and quality of lifein subclinical hypothyroidism: randomized, crossovertrial. J Clin Endocrinol Metab 92: 1715–1723.

26. Escobar-Morreale HF, Botella-Carretero JI, Gomez-BuenoM, Galan JM, Barrios V, et al. (2005) Thyroid hormonereplacement therapy in primary hypothyroidism: arandomized trial comparing L-thyroxine plus liothyroninewith L-thyroxine alone. Ann Intern Med 142: 412–424.

27. Monzani F, Caraccio N, Kozakowa M, Dardano A,Vittone F, et al. (2004) Effect of levothyroxine replace‐ment on lipid profile and intima-media thickness insubclinical hypothyroidism: a double-blind, placebo-controlled study. J Clin Endocrinol Metab 89: 2099–2106.

28. Meier C, Beat M, Guglielmetti M, Christ-Crain M, StaubJJ, et al. (2004) Restoration of euthyroidism acceleratesbone turnover in patients with subclinical hypothyroidism:a randomized controlled trial. Osteoporos Int 15: 209–216.

29. Clyde PW, Harari AE, Getka EJ, Shakir KM (2003) Com‐bined levothyroxine plus liothyronine compared withlevothyroxine alone in primary hypothyroidism: arandomized controlled trial. JAMA 290: 2952–2958.

30. Sawka AM, Gerstein HC, Marriott MJ, MacQueen GM,Joffe RT (2003) Does a combination regimen of thyroxine(T4) and 3,5,3'-triiodothyronine improve depressive symp‐toms better than T4 alone in patients with hypothyroid‐ism? Results of a double-blind, randomized, controlledtrial. J Clin Endocrinol Metab 88: 4551–4555.

31. Christ-Crain M, Meier C, Guglielmetti M, Huber PR,Riesen W, et al. (2003) Elevated C-reactive protein andhomocysteine values: cardiovascular risk factors inhypothyroidism? A cross-sectional and a double-blind,placebo-controlled trial. Atherosclerosis 166: 379–386.

32. Caraccio N, Ferrannini E, Monzani F (2002) Lipoproteinprofile in subclinical hypothyroidism: response to levo‐thyroxine replacement, a randomized placebo-controlledstudy. J Clin Endocrinol Metab 87: 1533–1538.


33. Meier C, Staub JJ, Roth CB, Guglielmetti M, Kunz M, etal. (2001) TSH-controlled L-thyroxine therapy reducescholesterol levels and clinical symptoms in subclinicalhypothyroidism: a double blind, placebo-controlled trial(Basel Thyroid Study). J Clin Endocrinol Metab 86:4860–4866.

34. Monzani F, Di Bello V, Caraccio N, Bertini A, Giorgi D,et al. (2001) Effect of levothyroxine on cardiac functionand structure in subclinical hypothyroidism: a doubleblind, placebo-controlled study. J Clin Endocrinol Metab86: 1110–1115.

35. Li X, Wang Y, Guan Q, Zhao J, Gao L (2017) The lipid-lowering effect of levothyroxine in patients with subclini‐cal hypothyroidism: a systematic review and meta-analysis of randomized controlled trials. Clin Endocrinol(Oxf) 87: 1–9.

36. Yamamoto JM, Benham JL, Nerenberg KA, Donovan LE(2018) Impact of levothyroxine therapy on obstetric, neo‐natal and childhood outcomes in women with subclinicalhypothyroidism diagnosed in pregnancy: a systematicreview and meta-analysis of randomised controlled trials.BMJ Open 8: e022837.

732 Chen et al.

hypothyroidism: a systematic review and meta-analysis

Documents