Thyroid Function in Adults with Prader–Willi Syndrome; a Cohort Study and Literature Review

Prader–Willi syndrome (PWS) is a complex genetic syndrome combining hypotonia, hyperphagia, a PWS-specific neurocognitive phenotype, and pituitary hormone deficiencies, including hypothyroidism. The low muscle mass associated with PWS causes a low energy expenditure due to a low basal metabolic rate. Combined with increased energy intake due to hyperphagia, this results in a high risk of obesity and associated cardiovascular disease. To reduce the high mortality in PWS (3% yearly), exercise is extremely important. As hypothyroidism can impair exercise tolerance, early detection is crucial. We performed a literature search for articles on hypothyroidism in PWS, measured thyroid hormone (TH) levels in 122 adults with PWS, and performed a medical file search for medication use. Hypothyroidism (low free thyroxin) was present in 17%, and often central in origin (80%). Triiodothyronine levels were lower in patients who used psychotropic drugs, while other TH levels were similar. One in six patients in our cohort of adults with PWS had hypothyroidism, which is more than in non-PWS adults (3%). We recommend yearly screening of free thyroxin and thyroid-stimulating hormone levels to avoid the negative effects of untreated hypothyroidism on basal metabolic rate, body mass index, and cardiovascular risk. Additionally, we recommend measuring TH concentrations 3–4 months after the start of growth hormone treatment.


Materials and Methods
Ethical review and approval were waived for this study by the Medical Ethics Committee of the Erasmus University Medical Center, Rotterdam, the Netherlands. In this retrospective study, we reviewed the medical files of adults that visited the multidisciplinary outpatient clinic of our PWS reference center in the Erasmus University Medical Center, between January 2015 and December 2020, and underwent our routine systematic health screening. This systematic screening consists of a structured interview, a complete physical examination, a medical questionnaire, a review of the medical file including medication use, biochemical measurements and, if indicated and feasible, additional tests, as described previously (see [18]).
During the visit, blood samples were taken for general medical screening, including evaluation of thyroid function (fT4, triiodothyronine (T3), TSH). The reference values in our center for TSH were 0.4-4.3 mU/L before 1 February 2019, and 0. 56-4.27 mU/L after that date. The reference values for fT4 were 11-25 pmol/L (Ortho Vitros ® assay, Vitros ECI Immunodiagnostic System; Ortho-Clinical Diagnostics, Rochester, MI, USA) before 12 April 2019, and 13.5-24.3 pmol/L after that date (Fuijrebio Lumipulse ® assay). Reference values in our center for T3 were 1.4-2.5 nmol/L before 12 April 2019, and 0.7-2.0 nmol/L after that date. TSH, fT4 and T3 measurements changed methods during the study, but they were calibrated similarly, as checked by external quality assessment schemes.
Overt hypothyroidism was defined as an fT4 concentration below the reference range. Central overt hypothyroidism was defined as an fT4 concentration below the reference range, with a TSH concentration below or within the reference range. Primary hypothyroidism was defined as an fT4 concentration below the reference range, with a TSH concentration above the reference range. Overt hyperthyroidism was defined as an fT4 concentration above the reference range. If patients used levothyroxine before visiting our reference center, the diagnosis of overt hypothyroidism was based on referral letters and/or laboratory measurements before the start of levothyroxine; in that case, the distinction between primary or central hypothyroidism was also based on referral letters or, if available, on the laboratory measurements before the start of levothyroxine compared to the local reference values.
Subclinical hypothyroidism was defined as a normal fT4 concentration, with a TSH concentration above the reference range, based on a single measurement. It is important to note that the diagnosis of subclinical hypothyroidism is less reliable in adults with PWS. In the general population, TSH can be affected by obesity [19][20][21][22]. Furthermore, hypothyroidism can be both primary and central in PWS. Taken together, this means that TSH and, therefore, the diagnosis of subclinical hypothyroidism, should be interpreted with caution.
We investigated the relationship between TH measurements and genotype, as this relationship is still largely unknown. As gender, age, BMI, and GH treatment are known to influence TH in the general population, we also investigated their effect on TH concentrations in our cohort of adults with PWS [23][24][25][26][27][28][29][30].

Literature Review
We performed a search on Embase, Medline, the Web of Science Core Collection, the Cochrane Central Register of Controlled Trials, and Google Scholar for articles that describe thyroid function and/or TH measurements in patients with PWS. The search was last updated on 22 July 2021. For the full search strategy, see Table S1.
Inclusion criteria were: original research articles that described the prevalence of thyroid abnormalities (including, but not limited to: central and primary hypothyroidism, hyperthyroidism, and subclinical hypothyroidism) or TH measurements (including, but not limited to: thyroxine (T4), fT4, T3, free T3 (fT3), reverse T3 (rT3), and TSH) in methylationpositive individuals with PWS. Exclusion criteria were: meeting reports, workshop summaries, conference abstracts, guidelines, articles that included 10 or fewer subjects with PWS, articles that were not available online, and articles that were not available in English. When the same population was described in multiple articles, the population with the most laboratory values or the largest population was included. When an article described thyroid function before and after the start of GH treatment, only data at baseline were included in the table. Authors were contacted to clarify data when needed.

Statistical Analysis
Statistical analysis was performed using R version 3.6.3. Descriptive statistics for continuous variables are reported as median (interquartile range (IQR)). For dichotomous variables we display the number of patients and the percentage of the total number of patients, n (%). We used a chi-squared test to compare the prevalence of hypothyroidism between males and females, between paternal deletion and mUPD, between patients who did and did not use GH treatment, and between patients who did and did not use psychotropic drugs. To investigate the relationship between hypothyroidism, and BMI and age, we used the Wilcoxon rank sum test. We also used the Wilcoxon rank sum test to investigate the relationships between gender, genotype, and use of GH treatment and psychotropic drugs on the one hand, and laboratory measurements (fT4, T3, and TSH) on the other hand. If there were ties, an exact calculation method was used. The Kendall rank correlation test was used to assess correlations between age and BMI on the one hand, and laboratory measurements (fT4, T3, and TSH) on the other hand. As this was an exploratory analysis, no correction for multiple testing was performed.

Hypo-and Hyperthyroidism
Hypothyroidism was present in 21 patients (17%). A total of 12 patients had central hypothyroidism, 3 patients had primary hypothyroidism, and in 6 patients it was unknown whether the hypothyroidism was central or primary. In 17 patients, the diagnosis of hypothyroidism was based on referral letters. TH concentrations were provided in the referral letter in five cases. Additionally, two patients were diagnosed during childhood by the pediatric endocrinology department at our reference center, and two patients were diagnosed during our systematic health screening. The median age at diagnosis of hypothyroidism was 18 years (IQR 13-27) (age at diagnosis was unknown in two patients). The median dose of levothyroxine in patients with hypothyroidism was 68.8 µg (IQR 50.0-100.0) daily. Additionally, one patient with very mild hypothyroidism did not receive any treatment.
Three patients (2%) had a normal fT4 concentration, with a TSH concentration above the reference range (subclinical hypothyroidism), while one patient was diagnosed with hyperthyroidism, treated with thiamazole. Although not statistically significant, hypothyroidism seemed to be more prevalent in females (23%) than in males (10%, p = 0.051). There was no relationship with genotype, age, BMI, or GH treatment (Tables 2-4).

Thyroid Hormone Levels
For the 97 patients without (subclinical) hypo-or hyperthyroidism, TH concentrations and the associations between TH concentrations and patient characteristics (gender, genotype, age, BMI, and use of GH treatment and psychotropic drugs) are shown in Tables 2-4. The median fT4 concentration was 16.5 pmol/L (IQR 14.3-18.5), the median T3 concentration was 1.9 nmol/L (IQR 1.7-2.3), and the median TSH concentration was 1.6 mU/L (IQR 1.1-2.3). T3 was significantly lower in older patients and in patients without current GH treatment, while fT4 and TSH levels were similar. Gender, genotype, and BMI were not significantly related to any of the TH measurements. To visualize the exact distribution of the TH concentrations, we show the T3, fT4, and TSH concentrations according to BMI in Figure 1A-C.

Psychotropic Drugs
We explored the relationship between TH concentrations and the use of psychotropic or antiepileptic drugs. Forty-nine patients used psychotropic drugs. Only two patients used antiepileptic medication, and both also used psychotropic drugs. Therefore, the relationship between the use of antiepileptic drugs and thyroid function was not further explored. Use of psychotropic drugs was not associated with hypothyroidism (Table 4). However, T3 was significantly lower in patients who used psychotropic drugs (median 1.7 nmol/L (IQR 1.6-2.0)) than in those who did not (median 2.1 nmol/L (IQR 1.7-2.3), p = 0.02). The mean age of patients using psychotropic drugs was 36 years, and the mean age of patients not using psychotropic drugs was 29 years. No associations for specific types of psychotropic drugs were found (Tables S2 and S3).

Literature Review
The results of our literature review are summarized in Tables 5-7. Only 5 studies reported thyroid function separately for adults, while the other 21 studies reported thyroid function in children (n = 12) or in mixed populations containing children and adults (n = 9). Paternal deletion was the most common genotype in all studies. The prevalence of hypothyroidism differed between 0% and 33% in most studies, with one study reporting a prevalence of 72%. However, this study only included 18 children with PWS who were only up to 2 years old and, in this study, thyroid axis dysfunction was defined as serum total T4 and/or serum fT4 levels below the 2.5th percentile of a reference population. The prevalence of hypothyroidism in studies that only included adults ranged between 5% and 13%. For mixed populations of both children and adults, this prevalence was between 0% and 26%. Only two studies that included adults reported whether the hypothyroidism was central or primary in origin [31,32]. Although central hypothyroidism was more prevalent (2% and 4%), primary hypothyroidism (0% and 2%) did also occur. One study reported on the prevalence of subclinical hypothyroidism in PWS, and showed a prevalence of 5% in children and 1% in adults. Additionally, 11 studies reported TSH, 5 total T4, 14 fT4, 5 total T3, 6 free T3, and 1 reverse T3 concentrations.

Clinical Recommendations
Based on the results of our cohort, the literature review, and our clinical expertise, we formulated practical clinical recommendations for the screening and treatment of hypothyroidism in adults with PWS ( Figure 2

Clinical Recommendations
Based on the results of our cohort, the literature review, and our clinical expertise, we formulated practical clinical recommendations for the screening and treatment of hypothyroidism in adults with PWS ( Figure 2).

Discussion
The prevalence of hypothyroidism detected in our cohort of 122 adults with PWS was 17%, compared to only 3% in non-PWS adults [56]. The risk of hypothyroidism was increased in all adults with PWS, regardless of gender, genotype, age, BMI, or use of GH treatment or psychotropic drugs.
Our prevalence of hypothyroidism was higher than that of most previous studies on hypothyroidism in adults and mixed cohorts of adults and children Table 3 (A, B). However, two large French studies both showed an even higher prevalence of hypothyroidism (26%) in patients with PWS of 16 years and older [47,48]. This indicates that, although the prevalence is variable, hypothyroidism is frequent in adults with PWS.
Compared to the general population, there are several aspects of PWS that increase the complexity of the diagnosis and treatment of hypothyroidism in these patients. An increased vulnerability to the effects of untreated hypothyroidism of the patients, diagnostic challenges, and altered TH metabolism make hypothyroidism a complex issue in adults with PWS.

Vulnerability of the Patients
The vulnerability of the patients makes the treatment of hypothyroidism an important topic. The common effects of hypothyroidism on the muscles and the brain can be especially harmful to adults with PWS, as they already have impaired exercise tolerance and brain function.

Exercise Intolerance and Cardiovascular Risk
Patients with PWS have a high risk of developing obesity due to hyperphagia and a low BMR inherent to the syndrome. Hypothyroidism can cause arthralgia, lethargy, exertion fatigue, shortness of breath, and muscle problems [12,57]; this makes it hard to exercise, and increases the risk of obesity. Hypothyroidism is also directly associated with a

Discussion
The prevalence of hypothyroidism detected in our cohort of 122 adults with PWS was 17%, compared to only 3% in non-PWS adults [56]. The risk of hypothyroidism was increased in all adults with PWS, regardless of gender, genotype, age, BMI, or use of GH treatment or psychotropic drugs.
Our prevalence of hypothyroidism was higher than that of most previous studies on hypothyroidism in adults and mixed cohorts of adults and children (Tables 5 and 6). However, two large French studies both showed an even higher prevalence of hypothyroidism (26%) in patients with PWS of 16 years and older [47,48]. This indicates that, although the prevalence is variable, hypothyroidism is frequent in adults with PWS.
Compared to the general population, there are several aspects of PWS that increase the complexity of the diagnosis and treatment of hypothyroidism in these patients. An increased vulnerability to the effects of untreated hypothyroidism of the patients, diagnostic challenges, and altered TH metabolism make hypothyroidism a complex issue in adults with PWS.

Vulnerability of the Patients
The vulnerability of the patients makes the treatment of hypothyroidism an important topic. The common effects of hypothyroidism on the muscles and the brain can be especially harmful to adults with PWS, as they already have impaired exercise tolerance and brain function.

Exercise Intolerance and Cardiovascular Risk
Patients with PWS have a high risk of developing obesity due to hyperphagia and a low BMR inherent to the syndrome. Hypothyroidism can cause arthralgia, lethargy, exertion fatigue, shortness of breath, and muscle problems [12,57]; this makes it hard to exercise, and increases the risk of obesity. Hypothyroidism is also directly associated with a decreased BMR, leading to a higher prevalence of obesity, which can further impair physical activity [58][59][60][61]. Obesity results in a high cardiovascular risk. Both indirect and direct cardiovascular effects of hypothyroidism make its early detection and treatment an important topic in this already vulnerable patient population [16].

Brain Function
Thyroid function is responsible for a variety of physiological processes in the adult brain [62]. Adult-onset hypothyroidism can affect both cognitive function and psychological health [63]. Hypothyroidism can impair cognition, concentration, information processing speed, memory, perceptual function, and executive function [64,65]. Treatment with levothyroxine can reverse these symptoms [66]. Furthermore, anxiety and depressive symptoms are frequently reported in patients with hypothyroidism. These symptoms also improve after treatment with levothyroxine, leading to an increased quality of life [67,68].
The increased vulnerability of the patients, combined with diagnostic challenges and altered thyroid hormone metabolism, make the diagnosis and treatment of hypothyroidism an important issue in adults with PWS.

Diagnostic Challenges
Diagnostic challenges include patients' delay, doctors' delay, and unreliability of TSH.

Patients' Delay
Due to the intellectual disability that is often present in PWS, patients are often unable to express their complaints. Especially when the symptoms associated with hypothyroidism are subtle (e.g., mild fatigue or muscle weakness, or a slightly changed bowel pattern), they will not be reported by the patients.

Doctors' Delay
In the general population, TH concentrations are usually measured when there is a clinical suspicion of hypothyroidism. Unexplained weight gain, reduced appetite, fatigue, and constipation are well-known clinical signs of hypothyroidism that will alert most physicians to measure TH concentrations [12]. However, in patients with PWS, these symptoms are often unreliable. Unexplained weight gain will often be attributed to hyperphagia. In addition, this constant craving for food will make it easy to miss a slight reduction in appetite. Fatigue due to hypothyroidism can be easily mistaken for daytime sleepiness due to reduced hypothalamic arousal, which is often present in PWS [2,18]. Lastly, constipation is already present in 40% of patients with PWS, and will not alert physicians to screen for hypothyroidism [69].

Unreliability of TSH
Apart from patients' and doctors' delay, there is another diagnostic challenge. Hypothyroidism can be both primary and central in PWS (Table 6). In our cohort, we also found that both central hypothyroidism (n = 12) and primary hypothyroidism (n = 3) were present. Serum TSH concentrations in patients with central hypothyroidism are often normal [70][71][72]. Furthermore, TSH can be affected by obesity [19][20][21][22]. Taken together, this means that TSH is less reliable in PWS. In our clinic, we have seen several examples of patients with untreated overt central hypothyroidism, which had been missed because the physician had only measured TSH and not fT4.
As symptoms of hypothyroidism are unreliable in patients with PWS, and hypothyroidism can be both primary and central, we recommend to screen for hypothyroidism by measuring serum TSH and fT4 concentrations every year.

Altered Thyroid Hormone Metabolism
Prescription of endocrine and non-endocrine medication may disturb TH metabolism. Likewise, altered levels of "hunger hormones" may affect TH concentrations. Examples of these TH metabolism-altering factors in adults with PWS include use of psychotropic drugs, growth hormone treatment, and disturbed leptin and ghrelin levels.

Psychotropic Drugs
Psychotropic drugs can cause a disturbed synthesis and metabolism of TH [73,74]. Compared to the general endocrine population, the population of adults with PWS is characterized by frequent use of psychotropic drugs [75], such as antipsychotics, anxiolytics, and antidepressants. Psychotropic drugs can influence the synthesis and metabolism of TH in a variety of ways, such as changing iodine capture or decreasing thyrotropin-releasing hormone (TRH) responsiveness. Psychotropic drugs can also cause an altered deiodination of T4 to T3 by stimulating deiodinase activity [73,74,76]. The iodothyronine deiodinases D1, D2, and D3 regulate the conversion from the prohormone T4 (which is produced by the thyroid gland, and biologically inactive) to the active hormone T3. This conversion takes place mainly in peripheral tissues [77][78][79]. In our population, 40% of the patients used psychotropic drugs. T3 was significantly lower in patients who used psychotropic drugs than in patients who did not. However, patients who used psychotropic drugs were older (mean age 36 years) compared to patients who did not use psychotropic drugs (mean age 29 years), which might have influenced the results [23,24]. We did not find an association between specific types of psychotropic drugs and TH measurements. This could be related to a lack of power, as these subgroups were small.

GH Treatment
GH treatment is often prescribed to children and young adults with PWS. In our population, one-third of the patients were treated with GH. Patients who used GH treatment had generally been receiving GH treatment for several years before visiting our outpatient clinic. It has been suggested that GH treatment enhances peripheral conversion of T4 to T3, resulting in decreased fT4 and increased T3 levels [80][81][82]. Several groups have studied the effect of GH treatment on TH levels, with contradictory results. Several studies showed that GH treatment does not induce hypothyroidism, but can unmask previously undiagnosed hypothyroidism in non-PWS individuals [83][84][85][86][87]. In children with PWS, fT4 concentrations decreased after the start of GH treatment, but remained within the low-normal range [25]. Two randomized controlled trials in adults with PWS showed no significant effect of GH treatment on fT4 and TSH [45,88], whereas one study showed increased T3 concentrations during GH treatment [45]. In our cohort, GH treatment was associated with higher T3 levels, while fT4 and TSH concentrations were similar for patients with and without GH treatment. To prevent the potential negative effects of missing "unmasked" hypothyroidism, we recommend measuring TH concentrations 3-4 months after the start of GH treatment.

Leptin and Ghrelin
High leptin levels caused by obesity in patients with PWS can lead to an increased conversion of T4 to T3 [89]. This mechanism might be partially responsible for the relatively low fT4 levels found in PWS patients [25,90]. Altered ghrelin levels in PWS [91,92] may impair the activity of the hypothalamic-pituitary-thyroid axis [93], which might increase the prevalence of hypothyroidism.

Strengths and Limitations
As with any study, our study has several strengths and limitations. One strength of our study is the population size, considering the rareness of the disease. Furthermore, we are the first to not only describe the prevalence of hypothyroidism and TH concentrations, but also the relationship with medication use. One of our limitations is that this was a retrospective study and, therefore, we had many missing values for T3, as this was not routinely measured in all patients. Another limitation is that thyroid peroxidase (TPO) antibodies were not measured; therefore, we were not able to distinguish autoimmune thyroid diseases. Furthermore, we did not measure thyroxine-binding globulin (TGB) and, therefore, we do not know whether free T3 levels were disturbed in our cohort. The last limitation is that laboratory measurements before the start of levothyroxine treatment were not available in 12 patients with hypothyroidism. In these cases we had to rely on referral letters that mentioned whether the patient had central or primary hypothyroidism.

Conclusions
In conclusion, one in six patients in our cohort of 122 adults with PWS had hypothyroidism, which is more frequent than in non-PWS adults. Hypothyroidism is often central in origin, and it is therefore important to measure not only TSH, but also fT4. We recommend yearly screening of fT4 and TSH to prevent the negative effects of untreated hypothyroidism on BMR, BMI, cardiovascular risk, and brain function. Additionally, we recommend measuring thyroid hormone concentrations 3-4 months after the start of GH treatment.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/jcm10173804/s1: Table S1: Full search strategy (Embase); Table S2: Prevalence of hypothyroidism and thyroid hormone levels in relation to use of psychotropic drugs (Part 1); Table S3: Prevalence of hypothyroidism and thyroid hormone levels in relation to use of psychotropic drugs (Part 2); Figure S1: Scatterplot of T3, fT4, and TSH in relation to BMI for patients who use psychotropic drugs versus patients who do not; Figure S2: Scatterplot of T3, fT4, and TSH in relation to BMI for patients who currently use growth hormone treatment versus patients who do not. Informed Consent Statement: Informed consent was obtained from subjects involved in the study, or anonymized patient data were collected.

Data Availability Statement:
The datasets generated during and/or analyzed during the current study are not publicly available, in order to protect the privacy of the patients participating in this study. As Prader-Willi syndrome is a rare syndrome, individual patient data could be traced back to the individual.