European Journal of Internal Medicine 126 (2024) 49–55

Contents lists available at ScienceDirect

European Journal of Internal Medicine

journal homepage: www.elsevier.com/locate/ejim

Original Article

The interplay between subclinical hypothyroidism and poor sleep quality: A

systematic review
Marsida Teliti a, b, Francesco Fanfulla c, Laura Croce a, b, Francesca Coperchini a,
Mario Rotondi a, b, *
a
Department of Internal Medicine and Therapeutics, University of Pavia, 27100, Italy
b
Unit of Endocrinology and Metabolism, Laboratory for Endocrine Disruptors, Istituti Clinici Scientifici Maugeri IRCCS, Via S. Maugeri 4, Pavia 27100, Italy
c
Unit of Respiratory Function and Sleep Medicine, Istituti Clinici Scientifici Maugeri IRCCS, Pavia 27100, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: Background: The relationship between subclinical hypothyroidism (SHYPO) and sleep disturbances is still poorly
Subclinical hypothyroidism investigated. This systematic review aims to critically appraise the existing literature to provide more insights in
Sleep quality understanding whether SHYPO favors sleep disturbances or it is the sleep disturbance per se that affects the
Sleep duration
hypothalamus-pituitary-thyroid axis regulation.
Systematic review
Sleep measures
Methods: Original studies on sleep quality and duration in patients with SHYPO were searched in the PubMed/
MEDLINE, Embase, Web of Science and Scopus databases. Two reviewers independently screened articles for
inclusion, extracted data, and assessed the quality of included studies.
Results: Eight studies, including 2916 patients with SHYPO and 18,574 healthy controls, were retrieved. An
overall agreement (7 out of 8 studies), about a positive correlation between decreased sleep quality and/or
duration and SHYPO was observed. Five studies investigated sleep quality through self-reported surveys; only
two studies explored both subjective and objective assessment of sleep quality with actigraphy (n = 1) or pol­
ysomnography (n = 1); finally, one study assessed subjective evaluation of sleep quality through a single question
regarding the number of sleeping hours. A high level of heterogeneity among studies was manifest due to dif­
ferences in population source, sleep measure assessment and criteria for diagnosing SHYPO.
Discussion: Overall, the existing literature data suggest a link between SHYPO and sleep disturbances, but further
studies on larger populations of patients with homogeneous study designs and outcomes are warranted.

1. Introduction intra-individual TSH levels are stable during daytime and begin to rise
and reach maximum levels during the early part of the night.
Sleep and the hypothalamic-pituitary-thyroid (HPT) axis show a Throughout the sleep period, TSH levels progressively decline, to return
mutual interdependence in normal physiological functioning [1]. While to daytime values shortly after morning awakening. While the evening
the interconnection between overt hypo- and hyperthyroidism and sleep increase is considered to reflect a circadian control, the nocturnal
disorders is well documented, there is a lack of substantiating evidence decrease during the sleep period seems to result from an inhibitory ac­
when mild thyroid failure is taken into account. Subclinical hypothy­ tion of sleep on TSH release. Indeed, the nocturnal decrease is not
roidism (SHYPO), defined by an elevated thyroid stimulating hormone observed in absence of sleep. Compared to a fully rested condition, in
(TSH) with normal serum level of thyroid hormones, is a growing public healthy young male volunteers one night of total sleep deprivation was
health concern and affects up to 10 % of the general population, espe­ associated with a near two-fold rise in nocturnal TSH levels and a
cially women. On the other hand, sleep quality is one of the most striking increase in free thyroxine index. Subsequent daytime recovery
common factors related to poor health outcomes [2]. Normal sleep sleep did not reduce TSH below daytime levels [4]. As sleep deprivation
promotes the regulation and release of thyroid hormones. Indeed, TSH continues, an overall decrease in TSH secretion has been suggested by
secretion exhibits a single sinusoid daily rhythmicity and in turn impacts some studies, presumably because of negative-feedback effects from
the quality and length of time asleep [3]. In healthy young adults, slowly rising concentrations of thyroid hormones. Whether prolonged

* Corresponding author.
E-mail address: [email protected] (M. Rotondi).

https://doi.org/10.1016/j.ejim.2024.03.013
Received 14 November 2023; Received in revised form 22 February 2024; Accepted 11 March 2024
Available online 28 March 2024
0953-6205/© 2024 The Author(s). Published by Elsevier B.V. on behalf of European Federation of Internal Medicine. This is an open access article under the CC BY
license (http://creativecommons.org/licenses/by/4.0/).
M. Teliti et al. European Journal of Internal Medicine 126 (2024) 49–55

sleep deprivation has an inhibitory effect on TSH or are the increasing 2.3. Data extraction
thyroid hormone levels to exert an inhibitory effect on sleep remains
controversial [5]. To further complicate the issue, Kessler et al. reported For each included study, information was extracted concerning
a modest but statistically significant decline in circulating TSH and reference data (authors, journal, year of publication, type of study and
free-thyroxine (fT4) concentrations, in females but not in males, after country of origin). Data regarding sample size, age, gender, body mass
two weeks of recurrent restricted sleep suggesting gender-based differ­ index (BMI), sleep health assessment and biochemical evaluation of
ences in the impact of sleep on the HPT axis activity. It was speculated thyroid function parameters were also extracted.
that reduced thyroid hormones in response to long term sleep restriction
may reflect an adaptation to counteract excessive catabolism, common 2.4. Quality assessment
in chronic conditions [6]. As concerns circadian disruption, a recent
meta-analysis revealed higher serum TSH and fT4 levels in night shift The quality of included studies was assessed using the National In­
workers than non-night shift workers [7]. However, careful evaluation is stitutes of Health (NIH) tool for Observational Cohort Studies and Cross-
necessary, since this alteration could be caused by a phase shift of the Sectional Studies. All the studies were assigned a yes, no, or other to
TSH rhythm. Moreover, according to the results provided by this each of the 14 criteria outlined in the appraisal tool. Each item of
meta-analysis, it is clear that the effect of shift work on thyroid auto­ methodological quality was classified as yes, no, or not reported. Items
antibodies remains controversial. Despite a high prevalence of sleep rated as ’not applicable’ were not scored. Based on number of yes as
problems reported among patients with overt hypothyroidism, to our total score, studies were classified according to quality rating: Poor<50
knowledge, no previous systematic review has examined the relation­ %, Fair 50–75 % and good >75 %. (Table 1). Possible disagreements on
ship between SHYPO and sleep quality and/or duration. The literature the final score were resolved by consensus among the authors.
was systematically reviewed to retrieve the largest number of articles
reporting sleep quality and duration in patients diagnosed with SHYPO. 3. Results
This literature overview will mainly be directed to explore the
following three issues: 1) exploring size and quality of the literature 3.1. Retrieved articles
about sleep quality and/or duration in patients with SHYPO; 2) exam­
ining the literature evidence on the correlation between sleep quality Using the above search strategy, 70 records were initially found.
and/or duration and SHYPO; 2) investigating the possible variables that Among these, 62 were excluded because not fitting with the study aim
determine sleep impairment in patients with SHYPO. while the remaining 8 were included in the systematic review. Fig. 1
The final and specific aim of the present review will be to critically illustrates the flow of articles
revise the available data in order to provide an answer to the still un­
solved question of whether SHYPO impacts sleep quality or if sleep 3.2. General features of included articles
disturbances, in turn, induce an increase in TSH levels.
The 8 articles included in the systematic review contained a total
2. Materials and methods number of 2916 patients with SHYPO and 18,574 healthy controls. The
studies were published from June 2014 to October 2022 by authors from
2.1. Search strategy six countries: 1 study from Italy, 1 from United States of America, 1 from
Korea, 1 from Sweden, 3 form China and 1 from Turkey. One was a
A comprehensive computer literature search of the PubMed/MED­ single cross-sectional study, four were comparative cross-sectional
LINE, Embase, Web of Science and Scopus databases was conducted to studies, two were cohort studies, and 1 reported both comparative
find published articles on the topic of our review. The search algorithm cross-sectional and longitudinal findings. SHYPO was diagnosed by
was created based on combinations of specific terms: (“Sleep quality” measuring serum levels of TSH and FT4 in all studies, free triiodothy­
OR “Sleep duration”) AND (“subclinical hypothyroidism”). A beginning ronine (fT3) levels and thyroid autoantibodies were measured in 4. The
date limit was not used, and the search was updated until October 30th, majority of the studies investigated sleep quality through self-reported
2023. To identify additional studies and expand our search, the refer­ surveys such as the Pittsburgh Sleep Quality Index (PSQI, n = 6),
ences of the retrieved articles were also screened. The authors declare Epworth Sleepiness Scale (ESS, n = 1), NHP (Nottingham Health Profile)
they have no conflict of interest. This research did not receive any sleep subscale (n = 1); only two studies explored objective assessment of
specific grant from funding agencies in the public, commercial, or not- sleep quality with wrist actigraphy (n = 1) or polysomnography (PSG, n
for-profit sectors. The review protocol was previously registered on = 1); finally, 1 study assessed subjective evaluation of sleep quality
PROSPERO (registry number CRD42023461113). through single question [8–12]. Table 2 shows the key characteristics
and the main findings of the studies included.
2.2. Study selection
3.3. Main socio-demographic and clinical characteristics of study
Studies or subsets of studies that report data on the detection of sleep population (age, gender, BMI and thyroid autoantibodies)
quality and sleep duration in patients with SHYPO were eligible for
inclusion. The main exclusion criteria were a) articles not within the Sayilan et al. studied 103 early to middle-aged patients (mean age =
field of interest of this review; b) articles including subclinical hypo­ 44±14.12 years), mostly men, presenting on an outpatient setting from
thyroid patients previously diagnosed with obstructive sleep apnea two separate hospitals in Turkey [13]. A total of 287 middle-aged in­
(OSA) c) review articles, editorials, or comments; d) articles that did not dividuals, divided into an isolated TSH elevated group and a normal TSH
provide clear study characteristics or reports that had overlapping pa­ group, were recruited by Yan et al. [14]. A comparative cross-sectional
tient data; e) non-English language articles. No restriction on study and longitudinal study was conducted, and the inclusion criteria were
design was applied. Two authors (M.T., L.C.) independently reviewed patients aged 18–65 year with a BMI of 18–24 kg/m2. The serum level of
the titles and abstracts of the retrieved articles, applying the inclusion TSH, FT3, FT4, thyroid peroxidase antibodies (TPO Ab), and thyro­
and above-mentioned exclusion criteria. Then, the same two researchers globulin antibodies (Tg Ab) were examined. Subjects with TPO Ab > 68
independently reviewed the full text of the articles to determine their IU/ml were diagnosed with chronic autoimmune thyroiditis.
final inclusion. Meanwhile, in the study conducted by Wu et al., data were obtained
from the national epidemiological survey (Tide) focused on the preva­
lence of thyroid diseases and diabetes and the iodine nutrition status

50
M. Teliti et al. European Journal of Internal Medicine 126 (2024) 49–55

Table 1
Quality Assessment Tool of Observational Cohort and Cross-Sectional Studies.
Year Author 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Total Score Quality Rating

2022 Sayilan et al. Y Y NR Y Y N N N Y N Y NA N N 6/13 (46.1 %) Poor

2022 Yan et al. Y Y NR Y N Y Y Y Y Y Y NA N Y 10/13 (71.4 %) Fair
2021 Wu et al. Y Y NR Y N N N N Y N Y NA N Y 6/13 (46.1 %) Poor
2021 Ellegard et al. Y Y Y Y N N N N Y N Y NA N N 6/13 (46.1 %) Poor
2019 Kim et al. Y Y NR Y N N N Y Y N Y NA N Y 7/13 (53.8 %) Fair
2019 Song et al. Y Y NR Y N N N N Y N Y NA N Y 6/13 (46.1 %) Poor
2018 Andaloro et al. Y Y NR Y N N N N Y N Y NA N N 5/13 (38.4 %) Poor
2014 Akatsu et al. Y Y NR Y N N N N Y N Y NA N Y 6/13 (46.1 %) Poor

Quality of included studies was assessed using the National Institutes of Health (NIH) Quality Assessment tool for Observational Cohort and Cross-Sectional Studies (htt
ps://www.nhlbi.nih.gov/health-pro/guidelines/in-develop/cardiovascular-risk-reduction/tools/cohort). 1. Was the research question or objective in this paper
clearly stated? 2. Was the study population clearly specified and defined? 3. Was the participation rate of eligible persons at least 50 %? 4. Were all the subjects selected
or recruited from the same or similar populations (including the same time period)? Were inclusion and exclusion criteria for being in the study prespecified and
applied uniformly to all participants? 5. Was a sample size justification, power description, or variance and effect estimates provided? 6. For the analyses in this paper,
were the exposure(s) of interest measured prior to the outcome(s) being measured? 7. Was the timeframe sufficient so that one could reasonably expect to see an
association between exposure and outcome if it existed? 8. For exposures that can vary in amount or level, did the study examine different levels of the exposure as
related to the outcome (e.g., categories of exposure, or exposure measured as continuous variable)? 9. Were the exposure measures (independent variables) clearly
defined, valid, reliable, and implemented consistently across all study participants? 10. Was the exposure(s) assessed more than once over time? 11. Were the outcome
measures (dependent variables) clearly defined, valid, reliable, and implemented consistently across all study participants? 12. Were the outcome assessors blinded to
the exposure status of participants? 13. Was loss to follow-up after baseline 20 % or less? 14. Were key potential confounding variables measured and adjusted
statistically for their impact on the relationship between exposure(s) and outcome(s)?.
Total Score: Number of yes; NA, not applicable; NR, not reported; N, no; Y, yes.
Quality Rating: Poor <50 %, Fair 50–75 %, Good ≥75 %.

[15]. The 159 SHYPO patients selected were all adult residents, aged > ancillary study [20]. This prospective cohort study involved 682
18 years (mean age = 42.4 ± 15.8 years), approximately 50 % male, community-dwelling older men with complete thyroid function data at
from Han ethnicity and living in Fujian for at least 5 years. A total of 159 baseline, actigraphy measurements, and PSQI and ESS questionnaire
EU age and gender matched participants were enrolled as control group assessment. Over 90 % of the men were classified as euthyroid, 6 % were
by propensity score matching. Positive rate of TPO Ab and Tg Ab were classified as subclinical hypothyroid and 2 % were classified as sub­
significantly increased in patients with SHYPO compared with EU par­ clinical hyperthyroid. Compared to EU patients, men with SHYPO were
ticipants. Moreover, a trend towards higher BMI in patients with SHYPO older. No difference in BMI score was found among the three groups.
compared with EU subjects was found.
In the study by Ellegard et al., the analyses data were derived from
3.4. Overall findings on sleep assessment in SHYPO
the World Health Organization (WHO) MONItoring of trends and de­
terminants for Cardiovascular disease (MONICA) project [16]. Patients
In 2014, Akatsu et al., for the first time evaluated the association
initially enrolled in 1995 were invited for a follow-up and a
between subclinical thyroid diseases and sleep disturbances assessed
re-examination 13 years later. A comparison between subjects with an
both subjectively and objectively; 682 older men were recruited for the
unawareness of their newly detected elevated TSH and EU controls was
study, 15 were classified as subclinical hyperthyroid and 38 were clas­
performed. From data, subject with SHYPO, mostly female, were older
sified as subclinical hypothyroid. No statistically significant differences
(mean age = 66.4 ± 8.3 years vs mean age = 62.1 ± 9.6 years) and had
in sleep quality and duration were found between subclinical hypothy­
more frequently positive TPO Ab (47% vs 9 %) than EU patients. No
roid and EU men [20].
significant differences in BMI and gender were found.
Conversely, according to Andaloro et al., subclinical thyroid disease
Kim et al., investigated the association between sleep duration and
affects sleep quality negatively. Patients with subclinical thyroid disease
subclinical thyroid dysfunction using the sixth Korea National Health
exhibited a statistically significant higher PSQI score and higher apnea-
and Nutrition Examination Survey database [17]. The study cohort
hypopnea index (AHI) in comparison to EU subjects. However, the
consisted of 2097 short sleepers, 2514 normal sleepers and 334 long
prevalence of moderate to severe obstructive sleep apnea (OSA) was
sleepers. No difference in TPO Ab level was found. SHYPO was diag­
higher in patients with SHYPO in comparison to hyperthyroid and EU
nosed in 149 patients: 12 were long sleepers (4.04 %), 74 were short
participants [19].
sleepers (3.73 %) and 63 were normal sleepers (2.35 %). Long sleepers
By using PSQI questionnaire, Song et al., confirmed a significant
showed higher mean TSH, were younger and had lower BMI as
relationship between SHYPO and poor sleep quality in a large Chinese
compared with both short and normal sleepers.
population. Compared to the EU patients, subjects with SHYPO were
Song et colleagues enrolled individuals with unspecified inclusion
associated with poorer sleep, longer sleep latency and shorter sleep
and exclusion criteria who underwent routine check-ups at the Health
duration, after adjusting for confounding factors [18]. In the study by
Promotion Center of West China Hospital and completed the PSQI
Kim et al., a cohort of 2543 males and 2402 females were stratified by a
questionnaire. [18]. The SHYPO group had a higher average age (43.77
self- reported habitual sleep duration. The risk of SHYPO in short
±9.48 vs 42.95±8.85) and female population (55.3% vs 40 %) than the
sleepers was similar to the risk of in normal sleepers, while long sleepers
EU subjects. No significant difference in BMI was found.
showed a higher incidence of SHYPO than normal sleepers after
Andaloro et al., in a different clinical setting study, enrolled
adjusting for possible potential confounders [17].
consecutive patients with a diagnosis of SHYPO (n = 45) and subclinical
According to Ellegard et al., a subgroup of 30 subjects with newly
hyperthyroidism (SHYPE, n = 38) who had referred to a sleep clinic from
detected, and unaware of, elevated serum level of TSH presented higher
2016 to 2017. Forty EU patients were selected as control group. No
score in the NHP-subscales Sleep, compared to EU controls, suggesting a
significant differences in BMI score, age and gender were found among
poorer sleep quality [16]. The results were confirmed after adjustment
the three groups [19].
for age, sex and comorbidities. In a subsequent study, Wu et al. inves­
Lastly, Akatsu et al., evaluated subjects from the Osteoporotic Frac­
tigated the association between lifestyle and thyroid function in SHYPO
tures in Men (MrOS) study, which subsequently participated in an
by including sleep quality. PSQI scale scores were collected. Patients

51
M. Teliti et al. European Journal of Internal Medicine 126 (2024) 49–55

Fig. 1. PRISMA 2020 flow diagram of the study selection process.

with SHYPO were more likely to have an overall poor sleep quality when factors for SHYPO adjusted by possible confounding factors (age, sex,
compared with EU subjects [15]. New findings by Yan et al., confirmed BMI, systolic blood pressure, Tg Ab, total cholesterol, uric acid and
that patients with an isolated TSH elevation experience poor sleep (as Hemoglobin A1C) in two logistic regression models. Both models
assessed by PSQI), more frequently than that observed in the general showed that patients with poor overall sleep quality had an increased
population [14]. Sayilan et al. found a moderate reduction in sleep risk of SHYPO independently from Tg Ab positivity [15].
quality and a moderate to severe reduction in quality of life in SHYPO In addition to demographic variables, the factors that showed sig­
patients. In addition, the Authors found a negative correlation between nificant difference between groups (BMI, diabetes, alcohol consump­
PSQI score and mean cognitive function, social relationship, and quality tion, household income level, education, urine iodine creatinine ratio,
of life scores [13]. and working pattern). were considered by Kim et coworkers. After
adjusting for all possible confounding factors, long sleepers exhibited a
significantly increased risk of SHYPO in 2 out of 3 logistic regression
3.5. Dealing with confounding factors analysis. A subgroup analysis was performed in subjects who were
negative for TPO Ab. Surprisingly, the statistical significance of the
Potential confounding factors for poor sleep quality in patients with increased risk of SHYPO in long sleepers was more prominent [17].
SHYPO were assessed only in five studies. A regression analysis was performed also by Song et al. to determine
Yan et al., constructed a multiple logistic regression analysis model the association between SHYPO and poor sleep quality. After adjusting
showing that both poor sleep and Tg Ab positivity were positively for potential confounders in different models, SHYPO was associated
associated with isolated TSH elevation. These results were confirmed in with higher PSQI score, longer sleep latency and sleep duration < 5 h.
a subsequent subgroup analysis of subjects without chronic autoimmune When the Authors tried to explore the possible risk factors of poor sleep
thyroiditis (which in this study had been defined according to a value of quality in patients with SHYPO, the risk of poor sleep quality decreased
TPO Ab at least double as compared to the normal limit; i.e. TPO Ab < as age and BMI increased in female patients. Conversely, there was no
68 IU/ml) [14]. Wu et al., analyzed the risk factors and protective

52
M. Teliti et al. European Journal of Internal Medicine 126 (2024) 49–55

Table 2
Main characteristics of the included studies.
Author, Study design Comparison, Age, male/ BMI TPO Sleep measure SHYPO Relevant finding
country, n mean ± female Kg/m2 Ab/Tg assessment assesment
year [ref.] SD Ab

Sayilan, Cross-sectional SHYPO 103 44.46 78/25 NA NA PSQI, mean±SD Normal fT3 and Sleep quality of the patients
Turkey, ±14.12 7.11±3.75 fT4 value and followed up for SHYPO is at a
2022 high TSH moderate level
[13]
Yan, China, Cross-sectional, SHYPO 40.65 57/ Range 23.2%/ PSQI (% poor sleep > TSH > 4.20 mU/ Poor sleep was more prevalent
2022 longitudinal and 168EU 119 ±11.15 111 18-24 23.2% 5) L with fT3 and and apparent in patients with
[14] comparative 40.36 39/80 5.88/ 70.24% fT4 within the isolated TSH elevation than in
±10.49 6.72% 49.58% normal the general population.
laboratory ranges
Wu, China, Cross-sectional, SHYPO 159 42.4 81/78 mean 19.5%/ PSQI (% poor sleep TSH > 4.2 mU/L SHYPO patients had poorer
2021 comparative, EU 159 ±15.8 87/72 ± SD 26.4% ≥7) and normal fT4 overall and subjective sleep
[15] community- 40.6 23.9 8.8%/ 25.8% levels (12.0-22.0 quality, shorter sleep latency
based ±15.0 ±3.2 6.3% 15.1% pmol/L) and lower proportion of late
22.6 sleep on weekdays.
±3.1
Ellegard, Cohort study of SHYPO 30 66.4 7/23 mean 47%/ NHP sleep (mean±SD) TSH > 4.2 mU/L Subjects with SHYPO scored
Sweden, random EU 344 ±8.3 86/ ± SD NA 35.3±34.5 and normal fT4 worse HRQoL regarding NHP
2021 population 62.1 258 27.6 9%/NA 18.9±25.6 levels (12.0-22.0 Sleep
[16] sample ±9.6 ±4.5 pmol/L)
26.7
±4.7
Kim, Korea, Cross-sectional Short 46.71 1085/ mean 6.21%/ Self-reported TSH > 6.68 mU/ The risk of subclinical
2019 and comparative sleepers ±0.35 1012 ± SD NA questionnaire using L hypothyroidism in short
[17] 2097 24.22 the question “how sleepers was similar to that in
SHYPO 74 ±0.09 many hours of sleep do normal sleepers, while long
EU 1960 you usually get in a day sleepers showed a higher
SHYPE 63 on average? incidence of subclinical
Non-EU 137 hypothyroidism than normal
Normal 43.46 1307/ mean 5.95%/ Short sleepers (< 7h/ sleepers after adjusting for
Sleepers ±0.32 1207 ± SD NA day) possible confounding factors.
2514 23.67 Normal sleepers (7-8
SHYPO 63 ±0.07 h/day)
EU 2395 Long sleepers (>8h/
SHYPE 56 day)
Non-EU 119
Long 41.17 151/ mean 5.42%/
sleepers ±1.09 183 ± SD NA
334 23.45
SHYPO 12 ±0.28
EU 306
SHYPE 16
Non-EU 28
Song, Cross-sectional SHYPO 2224 43.77 994/ mean NA/NA PSQI, 1-week version TSH > 4.2 mU/L SHYPO was significant
China, population- EU 12622 ±9.48 1230 ± SD NA/NA (% poor sleep >5) and normal fT4 associated with increased risk
2019 based 42.95 7573/ 23.69 67.1% levels (12.0-22.0 of poor sleep quality
[18] ±8.85 5049 ±3.37 64.7% pmol/L) and fT3
23.83 (3.6-7.5 pmol/L)
±3.31
Andaloro, Cross sectional SHYPO 45 42.2 17/28 mean NA/NA PSQI (mean±SD) TSH > 5.0 mU/L Significant higher PSQI and
Italy, and comparative EU 40 ±12.82 14/26 ± SD NA/NA 11.1±4.3 and fT4 level AHI scores in subclinical
2018 SHYPE 38 40.7 13/25 27.3 NA/NA 4.4±2.3 within normal thyroid disease groups
[19] ±11.25 ±3.13 9.7±3.9 reference range
43.3 26.5
±11.93 ±3.32
25.7
±2.56
SHYPO 45 PSG (AHI event/h)
EU 40 23.5±10.6
SHYPE 38 7.8±5.1
19.9±8.8
Akatsu, Prospective SHYPO 38 75.3 38/0 mean NA/NA PSQI (% poor sleep TSH > 4.78 mU/ No statistically significant
USA, cohort EU 629 ±6.8 629/0 ± SD NA/NA >5; mean±SD) L and normal fT4 differences exist between
2014 SHYPE 15 72.7 15/0 27.2 NA/NA 50.0%; 6.0±3.6 levels (0.7-1.85 subclinical thyroid diseases
[20] ±5.5 ±3.3 46.1%; 5.6±3.2 ng/dL) and subjective or objective
74.1 27.6 40.0%; 6.9±5.1 sleep disturbance.
±4.8 ±3.7
27.2
±4.5
SHYPO 38 ESS (mean±SD)
EU 629 6.1±3.3
SHYPE 15 6.1±3.8
5.9±3.8
(continued on next page)

53
M. Teliti et al. European Journal of Internal Medicine 126 (2024) 49–55

Table 2 (continued )
Author, Study design Comparison, Age, male/ BMI TPO Sleep measure SHYPO Relevant finding
country, n mean ± female Kg/m2 Ab/Tg assessment assesment
year [ref.] SD Ab

SHYPO 38 Wrist Actigraphy

EU 629 Sleep Duration Hours
SHYPE 15 (mean±SD)
6.4±1.1
6.4±1.2
6.0±1.5

Abbreviations: SHYPO, subclinical hypothyroidism; SHYPE, subclinical hyperthyroidism; EU, euthyroidism; SD, standard deviation; BMI, body mass index; TPO Ab,
thyroid peroxidase antibodies; Tg Ab, thyroglobulin antibodies; PSQI, Pittsburgh Sleep Quality Index; EES, Epworth Sleepiness Scale; NHP, Nottingham Health Profile;
PSG, polysomnography; TSH, thyroid stimulating hormone; HRQoL, Health-Related Quality of Life, AHI, Apnea hypopnea index.

association between TSH levels and poor sleep quality in the SHYPO increased risk of raised TSH independently from Tg Ab positivity. In the
subject [18]. study by Kim et al., similar rates of patients with positive TPO Ab were
Three sets of models were run by Akatsu et al. to analyze the asso­ found according to sleep length. A slightly higher percentage of TPO Ab
ciation between TSH and each sleep outcome: unadjusted, age-clinic and Tg Ab positivity was reported between patients with TSH elevation
adjusted, and multivariate-adjusted. Known or suspected factors for versus normal TSH group (23.21% vs 5.88 % and 23.21% vs 6.72 % for
thyroid dysfunction and sleep parameters were examined for potential TPO Ab and Tg Ab, respectively) by Yan et al. In both of these afore­
confounding. The risk of poor sleep for the patients with SHYPO was not mentioned studies, when excluding subjects with chronic autoimmune
significantly different from that of the EU group in adjusted analyses thyroiditis, a higher incidence of poor sleep in subjects with TSH
[20]. elevation was observed. It should be remembered that although TPO Ab
is thought to be superior for identifying thyroid autoimmunity than Tg
4. Discussion Ab, both are risk factors for SHYPO [22,23]. Interestingly, Yan et al.,
longitudinally followed up a total of 105 subjects with TSH elevation
The relationship between subclinical thyroid dysfunction and inad­ after assessment of the sleep status. Throughout a 4–24 weeks of sur­
equate sleep remains a poorly investigated issue but still an intriguing veillance, they observed a significantly higher rate of TSH normalization
one. In view of the final aim of the present review, the results obtained in patients who had improved sleep as compared to those who still slept
from the systematic revision of the literature will be discussed in order to poorly. Taking together these results, it could be proposed that in these
provide more insights helpful for understanding whether SHYPO favors patients the TSH elevation may not be the consequence of chronic
sleep disturbances or is the sleep disturbance per se which affects the autoimmune thyroiditis suggesting the possibility that a raised serum
hypothalamus-pituitary-thyroid axis regulation. TSH might be the consequence rather than the cause of poor sleep
According to the results of this systematic review, a positive corre­ quality. The lack of systematic evaluation of thyroid autoantibodies
lation between decreased sleep quality and/or duration and SHYPO was together with the fact that even in some of those studies taking into
found in 7 studies out of 8 (corresponding to 2878 patients evaluated out account thyroid autoimmunity, Tg Ab were not measured, should be
of 2916). The only study in which such an association was not found is regarded as a potential confounder, if we want to ascertain a relation­
the one by Akatsu et al., who reported no significant association between ship between mild thyroid failure (which is in the vast majority of cases
poor sleep quality and neither subclinical nor overt hypothyroidism supported by chronic autoimmune thyroiditis) and other causes of
[20]. However, caution is needed in interpreting these findings and raised hyperthyrotropinemia, not always related to a clinical condition
drawing conclusions. This holds particularly relevant when considering of SHYPO. These might include the so called “isolated hyper­
that significant heterogeneity exists between 8 studies in terms of study thyrotropinemia” mainly observed in subjects with obesity and morbid
design, patients’ characteristics (e.g., sex, age, ethnicity, comorbidities), obesity in particular, as well as patients with an age-related raised TSH
criteria for diagnosing SHYPO and assessment of sleep quality, and [21,24]. In addition, in view of the well-known clustering of autoim­
adjustment for confounding variables influencing thyroid hormone mune diseases, it could be hypothesized that, at least some patients with
levels and sleep parameters. In this view, some specific features of the chronic autoimmune thyroiditis may also have other autoimmune con­
studies included in this review should be discussed. First of all, the study ditions. This aspect is not negligible in that non-thyroid autoimmune
by Akatsu et al., specifically enrolled only community-dwelling men diseases, in particular rheumatic disease, associated with pain. Of note,
aged ≥65 years, which, in view of the increasing prevalence of sleep several evidences indicated that pain experienced in a given autoim­
disturbances and of the physiological increase in TSH concentrations in mune disease is unique and it seems to impact the sleep quality of the
the elderly, might have underestimated the association [21]. Second and affected patients [25]. However, none of the studies included in this
more importantly, at difference with the other studies, the timing of systematic review has taken into account the co-existence of other
thyroid function parameters and sleep evaluation was not coincident. autoimmune disorders in patients with chronic autoimmune thyroiditis.
Indeed, thyroid hormone status was assessed 3.4 years before the sleep It is worth remembering that the potential confounder role of body
study, thus thyroid function may have changed by the time of the sleep weight status does not represent a negligible issue. Indeed, the fact that
evaluation. both poor sleep quality and raised serum levels of TSH levels are
A major limitation for the interpretation of the SHYPO/sleep quality observed at an increased prevalence in the obese population, as
relationship stems from the lack of systematic evaluation of thyroid compared to normal-weight subjects, was extensively reported [23,24].
autoantibodies. In fact, only two studies out of eight measured both Tg Although currently available studies evaluating the relationship be­
Ab and TPO Ab whereas four studies took into account only TPO Ab. In tween SHYPO and sleep quality have not always performed a correction
the study by Ellegard et al., patients with elevated serum TSH, randomly for BMI, further discussing this point seems worthwhile. The studies
selected from the general population, displayed higher prevalence of included in this review focused on adults with BMI ranging from 18
positive TPO Ab tests and higher score in the NHP-subscales Sleep, kg/m2 to 32.1 kg/m2. The study of Wu et al.. reported a trend towards
suggesting a poor sleep quality compared to EU controls. However, this higher BMI in patients with SHYPO compared with age and
relationship was irrespective of TPO Ab level. Similarly, Wu et al. gender-matched EU subjects [15]. Thus, the influence of sample selec­
demonstrated that patients with poor overall sleep quality had an tion due to the exclusion of unmatched subjects cannot be excluded. It is

54
M. Teliti et al. European Journal of Internal Medicine 126 (2024) 49–55

questionable whether poor sleep quality diagnosed in patients with References

SHYPO does not merely represent non-causal comorbidity. Whether this
represents a causal relationship or a precise pathologic mechanism re­ [1] Shekhar S, Hall JE, Klubo-Gwiezdzinska J. The hypothalamic–pituitary–thyroid
axis and sleep. Curr Opin Endocr Metab Res 2021;17:8–14. 10.1016/j.
mains to be established, indeed the two conditions may simply coexist coemr.2020.10.002.
by chance. [2] Banks S, Dinges DF. Behavioral and physiological consequences of sleep restriction.
A further point to be highlighted here is that in 4 out of 8 studies, the J Clin Sleep Med 2007;3(5):519–28.
[3] Copinschi G, Caufriez A. Sleep and hormonal changes in aging. Endocrinol Metab
assessment of sleep disturbances was performed during the COVID-19 Clin North Am 2013;42(2):371–89. 10.1016/j.ecl.2013.02.009.
pandemic. This may represent an additional confounding factor since [4] Ikegami K, Refetoff S, Van Cauter E, Yoshimura T. Interconnection between
Jahrami et al. [26] reported that four in every ten individuals experi­ circadian clocks and thyroid function. Nat Rev Endocrinol 2019;15(10):590–600.
10.1038/s41574-019-0237-z.
enced a certain degree of sleep disturbance in this period and that a wide [5] Pereira JC, Andersen ML. The role of thyroid hormone in sleep deprivation. Med
spectrum of thyroid disturbances were reported in patients with Hypotheses 2014;82(3):350–5. 10.1016/j.mehy.2014.01.003.
COVID-19 [27]. [6] Kessler L, Nedeltcheva A, Imperial J, Penev PD. Changes in Serum TSH and Free T4
during human sleep restriction. Sleep 2010;33(8):1115–8. 10.1093/sleep/
The last and, at least in our opinion, strongest argument that should
33.8.1115.
be considered when depicting the intriguing relationship between TSH [7] Luo J. Assessment of the impact of shift work on thyroid disorders: a systematic
and sleep quality is related to gender. Indeed, when taking into account review and meta-analysis. Sleep Breath 2023;27(2):703–8. 10.1007/s11325-022-
gender of the patients with SHYPO enrolled by studies included in the 02652-9.
[8] Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The pittsburgh sleep
present review, it appears that a total of 1465 females and 1272 male quality index: a new instrument for psychiatric practice and research. Psychiatry
patients were studied. This indicates a female-to-male ratio of 1.15 Res 1989. 10.1016/0165-1781(89)90047-4.
which would not appear in line with the well-known female gender [9] Lapin BR, Bena JF, Walia HK, Moul DE. The Epworth sleepiness scale: validation of
one-dimensional factor structure in a large clinical sample. J Clin Sleep Med 2018;
prevalence observed in studies performed in SHYPO patients. It should 14(08):1293–301. 10.5664/jcsm.7258.
be mentioned that the study by Akatsu et al. [19], enrolled only male (n [10] Hunt SM, McEwen J. The development of a subjective health indicator. Sociol
= 38) while no gender data are available for the SHYPO patients Health Illn 1980;2(3):231–46. 10.1111/1467-9566.ep11340686.
[11] Jean-Louis G, Kripke DF, Mason WJ, Elliott JA, Youngstedt SD. Sleep estimation
enrolled by Kim, et al. [17]. Chronic autoimmune thyroiditis accounts from wrist movement quantified by different actigraphic modalities. J Neurosci
for the vast majority of cases of SHYPO, being characterized by one of Methods 2001;105(2):185–91. 10.1016/S0165-0270(00)00364-2.
the highest female to male ratio among human autoimmune conditions [12] Jafari B, Mohsenin V. Polysomnography. Clin Chest Med 2010;31(2):287–97.
10.1016/j.ccm.2010.02.005.
[28]. [13] Sayılan S, Ok E, Çetinkaya E. The association between sleep and quality of life in
Thus, the very similar rate of female and male patients enrolled patients diagnosed with subclinical hypothyroidism. Ankara Med. J. 2022;22(3):
would suggest, that, at least for some studies, patients were recruited in 424–33. 10.5505/amj.2022.94830.
[14] Yan Y. Elevated thyroid-stimulating hormone levels are associated with poor sleep:
a non-endocrine clinical settings and/or that patients were not consec­
a cross-sectional and longitudinal study. Endocrine 2022;75(1):194–201. 10.1007/
utively recruited, leading to potential selection bias. s12020-021-02849-0.
In conclusion, the results of the observational studies included in this [15] Wu K. Lifestyle is associated with thyroid function in subclinical hypothyroidism: a
systematic review would indicate an association between SHYPO and cross-sectional study. BMC Endocr Disord 2021;21(1):112. 10.1186/s12902-021-
00772-z.
sleep disturbances. However, a high degree of heterogeneity among [16] Ellegård L, Krantz E, Trimpou P, Landin-Wilhelmsen K. Health-related quality of
studies is evident due to differences in study design, population source, life in hypothyroidism A population-based study, the WHO MONICA Project. Clin
sleep measure assessment, criteria for diagnosing SHYPO, and further Endocrinol (Oxf) 2021;95(1):197–208. 10.1111/cen.14448.
[17] Kim W. Association between sleep duration and subclinical thyroid dysfunction
potential confounders taken into account (i.e. lack of systematic evalu­ based on nationally representative data. J Clin Med 2019;8(11):2010. 10.3390/
ation of thyroid antibodies and autoimmune conditions, role of gender jcm8112010.
differences and body weight status, and impact of COVID-19 pandemic). [18] Song L. The association between subclinical hypothyroidism and sleep quality: a
population-based study. Risk Manag Healthc Policy 2019;12:369–74. 10.2147/
Thus, the interplay between SHYPO-sleep quality/duration should be RMHP.S234552.
interpreted with caution to avoid the risk of overestimating such a [19] Andaloro C, La Mantia I. Subclinical thyroid disease and obstructive sleep apnea: a
clinical association. As a result, based on the currently available data, it focus on their relationship. Minerva Respiratory Med 2018;57(2). 10.23736/
S0026-4954.18.01813-8.
remains difficult understanding whether SHYPO can impair sleep [20] Akatsu H. Association between thyroid function and objective and subjective sleep
quality and duration, or if the opposite occurs—that is, if the sleep quality in older men: the osteoporotic fractures in men (MrOS) study. Endocrine
disturbance per se can promote an increase in serum TSH in the absence Pract. 2014;20(6):576–86. 10.4158/EP13282.OR.
[21] Yamada S. The impact of age- and sex-specific reference ranges for serum
of thyroid dysfunction.
thyrotropin and free thyroxine on the diagnosis of subclinical thyroid dysfunction:
To exhaustively address the above point, future studies, in particular a multicenter study from Japan. Thyroid 2023;33(4):428–39. 10.1089/
longitudinal prospective studies, experimental models as well as inter­ thy.2022.0567.
vention trials appear mandatory, in order to evaluate: [22] McLachlan SM, Rapoport B. Why measure thyroglobulin autoantibodies rather
than thyroid peroxidase autoantibodies? Thyroid 2004;14(7):510–20. 10.1089/
1050725041517057.
1) the effect of restoration of euthyroidism after levothyroxine [23] Fierabracci P. Possible added value of thyroglobulin antibody (TgAb) testing in the
replacement therapy on sleep quality in patients with SHYPO; evaluation of thyroidal status of subjects with overweight or obesity. J Endocrinol
Invest 2022;45(11):2077–84. 10.1007/s40618-022-01839-x.
2) the effect of adequate sleep restoration in patients with sleep dis­ [24] Santini F. Mechanisms in endocrinology: the crosstalk between thyroid gland and
turbances on serum thyroid function parameters. adipose tissue: signal integration in health and disease. Eur J Endocrinol 2014;171
(4):R137–52. 10.1530/EJE-14-0067.
[25] Mifflin KA, Kerr BJ. Pain in autoimmune disorders. J Neurosci Res 2016;95(6):
The issue appears of potential clinical relevance and future research 1282–94. 10.1002/jnr.23844.
seems worthwhile for a personalized therapeutic approach in patients [26] Jahrami HA. Sleep disturbances during the COVID-19 pandemic: a systematic
with sleep disturbances. review, meta-analysis, and meta-regression. Sleep Med Rev 2022;62. 10.1016/j.
smrv.2022.101591.
[27] Trimboli P. Thyroid sequelae of COVID-19: a systematic review of reviews. Rev
Endocr Metab Disord 2021;22(2):485–91. 10.1007/s11154-021-09653-1.
[28] Hennessey JV, Espaillat R. Subclinical hypothyroidism: a historical view and
shifting prevalence. Int J Clin Pract 2015;69(7):771–82. 10.1111/ijcp.12619.

55