Low awareness and under-diagnosis of hypothyroidism

Abstract

Hypothyroidism is a common condition with a prevalence that varies according to local dietary iodine availability, gender and age. The symptoms of hypothyroidism are generally nonspecific, with considerable overlap with other conditions and with the consequences for the health of advancing age. These symptoms are not useful for diagnosing hypothyroidism and a thyroid function test is required for a firm clinical diagnosis. Lack of knowledge and understanding of hypothyroidism, and a tendency for many people to attribute the symptoms of hypothyroidism to other causes have led to substantial unawareness and often late diagnosis of hypothyroidism. Large observational studies and meta-analyses have shown that about 4–7% of community-derived populations in the USA and Europe have undiagnosed hypothyroidism. About four cases in five of these are subclinical hypothyroidism, with the remainder being overt hypothyroidism. The prevalence of undiagnosed hypothyroidism is higher in older subjects, in women, and some ethnic groups, consistent with diagnosed disease. More research is needed to quantify the clinical burden of undiagnosed hypothyroidism around the world, with educational efforts aimed at the public and healthcare professionals aimed at identifying and managing these individuals.

Keywords:

• Hypothyroidism
• levothyroxine
• underdiagnosis
• thyroid stimulating hormone
• subclinical hypothyroidism

Introduction

A diagnosis of primary hypothyroidism is common, although reports of the prevalence of the condition are variable, due in part to regional variations in the availability of iodine from the diet¹. For example, the Colorado Health Study from the USA reported some 20 years ago that about 6% of its population of >25,000 community-dwelling subjects were taking thyroid medications². A meta-analysis reported a lower prevalence of hypothyroidism of about 3% in Europe³. Hypothyroidism is also common in India, with a reported prevalence of 11%⁴. The prevalence of hypothyroidism increases with age: cross sectional data from the US population-based Atherosclerosis Risk In Communities (ARIC) cohort found a prevalence of hypothyroidism of 24% among older subjects (aged ≥65 years)⁵.

Undertreatment of hypothyroidism, identified by persistent elevations of thyroid-stimulating hormone (TSH, thyrotropin) beyond its reference range, has been associated with adverse long-term outcomes, including increased risk of major adverse cardiovascular events, in large observational studies^6–10. We should remember that thyroid biomarkers are variable within and between subjects^11,12 and that observational studies are by their nature prone to confounding. Nevertheless, given the large numbers of people with hypothyroidism worldwide, even modest underdiagnosis (and thus undertreatment) would leave a substantial number of individuals at risk of such adverse long-term outcomes⁹. In this review, we have explored the magnitude of the problem of unawareness and underdiagnosis of hypothyroid disease.

Unawareness of hypothyroidism: scale of the problem

The ARIC study included cross-sectional analysis of thyroid hormones in 5,392 elderly (≥65 y), community-based individuals in the USA, using ARIC-derived cut-offs for thyroid hormone levels (see below for more discussion on the importance of defining cut-offs)⁵. Of these, 16.9% had a history of treatment for hypothyroidism, while biochemical evidence of subclinical and overt hypothyroidism was present in 6.1% and 0.8%, respectively. Within the ARIC dataset, the prevalence of hypothyroidism increased with age in treated and untreated groups, was higher in women versus men, with less subclinical or treated hypothyroidism (but not overt hypothyroidism) among Black versus White subjects. It should be noted that this study did not present information on prior diagnoses of hypothyroidism, so we cannot know what proportion of the previously untreated group were, both diagnosed and untreated.

The National Health and Nutrition Examination Survey (NHANES), a nationally representative cohort in the USA, also used its own reference population to define a cut-off for TSH for the diagnosis of hypothyroidism of >4.5 mIU/L¹³. This was applied to a population of 16,533 US citizens with thyroid hormone measurements but free of (self-reported) prior thyroid disease, among whom 4.1% had biochemical evidence of unsuspected hypothyroidism. This study did not evaluate a higher TSH cut-off to define overt hypothyroidism. The prevalence of any hypothyroidism, as defined, was higher in older and female subjects, and lower in non-white subjects, as in the study described above.

A systematic review and meta-analysis of 21 analyses published in 20 reports explored the prevalence of undiagnosed hypothyroidism in an overall population of 228,613 subjects in Europe¹⁴. The total prevalence of undiagnosed hypothyroidism was 4.70%, comprised of 4.11% and 0.65% with subclinical and overt hypothyroidism, respectively. Differences in prevalence in subgroups defined by age and gender were consistent with the large individual analyses described above (Figure 1). The prevalence of undiagnosed overt hypothyroidism in men was very low (0.06%) in this analysis. An earlier (2014) meta-analysis of 17 studies reported a similar prevalence of undiagnosed hypothyroidism in Europe of 4.9%, of which 4 cases in 5 were of the subclinical form³.

Figure 1. Prevalence of undiagnosed hypertension from a meta-analysis. Drawn from data presented in reference [14].

Undiagnosed hypothyroidism may be associated with other medical conditions. A cross-sectional study of 1160 subjects aged ≥60 years found a slightly higher prevalence of undiagnosed hypothyroidism, using age-appropriate reference intervals, in subjects with (6.4%) versus without (5.1%) diabetes¹⁵. In another study, the prevalence of undiagnosed subclinical hypothyroidism in a population of 82 overweight or obese subjects attending a sleep clinic for sleep-disordered breathing was 11.5%¹⁶. A high prevalence of subclinical hypothyroidism (9.7% in men and 14.6% in women) in elderly residents of a nursing home in the USA suggested potential benefit from screening this population for undiagnosed hypothyroidism¹⁷. Finally, a survey of emergency hospital admissions in an iodine-replete region of Taiwan found that 0.1% had overt hypothyroidism, most commonly associated with drugs (23%), autoimmune thyroiditis (21%) or previous surgery or irradiation of the thyroid gland (21%)¹⁸.

A study in the USA explored awareness of thyroid disease among NHANES participants between 1999 and 2004, focusing on subjects with self-reported receipt of thyroid hormone medication (levothyroxine or liothyronine/triiodothyronine)¹⁹. For the 2011–2014 cohort, 11% of these subjects stated they were unaware of having a thyroid condition, despite receiving this treatment. The demographic subgroups with the least awareness were men (24% unaware), subjects aged ≥65 years (16% unaware), non-Hispanic Black subjects (18% unaware), and subjects with less than high school education (also 18% unaware). It is important to remember that a patient with a biochemically adverse thyroid hormone profile in the absence of symptoms of hypothyroidism may not consider themselves to have a thyroid condition currently, while being aware that one was diagnosed at some point in the past, and surveys should be designed to elicit this information. Overall, the current evidence base suggest that low awareness of thyroid disease apparently continues to be an important barrier to optimal care for many patients even following the diagnosis of hypothyroidism.

What factors may account for widespread unawareness of hypothyroidism?

Nonspecific symptoms

People with hypothyroidism present typically with a range of relatively nonspecific symptoms, such as, fatigue, feeling cold, weight gain, constipation, low mood, slowed cognition, muscle aches/cramps, weakness, muscle cramps, dry skin, brittle hair and nails, diminished libido, carpal tunnel syndrome, or dysmenorrhea²⁰. Some of these symptoms are more prevalent in people with overt or subclinical hypothyroidism, compared with euthyroid subjects; however, the symptoms are found commonly in hypothyroid and euthyroid populations, and their presence or absence is not a reliable predictor of thyroid status^2,21. The symptoms reminiscent of hypothyroidism are similar to those of advancing age²², and persist in some individuals even after optimization of the TSH level with levothyroxine²³. Inevitably, many people will persevere with these nonspecific symptoms, perhaps attributing them to other causes, without discussing them with a healthcare professional. This may be especially true for subclinical hypothyroidism, where the severity of thyroid-related symptoms is likely to be lower than in an individual with overt clinical hypothyroidism.²

The diagnosis of hypothyroid disease is biochemical in nature and made according to a measurement of TSH that is higher than a reference range of TSH values derived from a euthyroid population (typically 0.4–4 mIU/L). Awareness by the patient of the need to discuss their symptoms with a healthcare professional, and awareness of the healthcare professional of the need to prescribe a thyroid function test are both needed to achieve a diagnosis of hypothyroidism, according to current guidelines²⁴. Once the patient does present with these symptoms, of if they are disclosed during a clinical consultation for another reason, it is important that physicians recognize a constellation of symptoms that are suggestive of hypothyroidism and investigate appropriately²⁴. Recent observations that clinical hypothyroid-like symptoms may correlate more reliably with thyroid hormones other than TSH (especially free T4 or triiodothyronine) may inform guidelines in the future and contribute to better identification of people who have (and importantly, do not have) hypothyroidism^19,25,26

Conversely, patients with thyroid dysfunction are more likely to report thyroid-related symptoms, such as neuropsychological disturbances or fatigue, once they are aware of their thyroid condition (reviewed elsewhere)²⁷. This phenomenon appears to be independent of the severity of thyroid dysfunction, as indicated by the level of TSH, and patients may seek to implicate symptoms arising from other causes to their known thyroid dysfunction²⁷. Patients’ self-reported health-related quality of life may decrease as a result of a diagnosis of hypothyroidism, or other noncommunicable diseases, per se²⁸.

Diagnostic issues

The diagnosis of overt hypothyroidism, where the TSH level is markedly elevated (>10 mIU/L), is straightforward once the patient has presented and undergone thyroid function testing. Measurement of smaller TSH elevations above the reference range is required for the diagnosis of subclinical hypothyroidism. Reference ranges for thyroid tests are assay-specific²⁹, and vary according to region, particularly with regard to the local level of iodine depletion, and the success of strategies to achieve iodine repletion³⁰. It has been noted recently that the use of different assays with different reference ranges might lead to different subsequent pathways for diagnosis and management for individual patients with suspected hypothyroid disease³¹. The use of age-specific reference ranges is important for optimizing the diagnosis of hypothyroidism, particularly for elderly patients¹⁵.

Interestingly, the analysis from the ARIC study, described above, derived its own diagnostic cut-offs for TSH from a healthy subset of the ARIC population. The presented prevalence values for subclinical or overt hypothyroidism in patients not receiving thyroid hormone replacement therapy were lower using these cut-offs compared with values defined by the cut-offs provided by the assay manufacturer (6.1% vs. 10.8% for subclinical hypothyroidism, and 0.8% vs. 2.2% for overt hypothyroidism). This study therefore provides an example of how the use of different reference populations can influence the diagnosis of thyroid dysfunction. These differences in prevalence rates would lead to substantial differences in the numbers of people with a diagnosis of hypothyroidism when extrapolated across a population.

Lack of follow-up of abnormal test results was identified as a cause of failure to diagnose overt hypothyroidism in a center in the USA. Here, identification of patients with elevated TSH who were not receiving thyroid hormone replacement revealed a subpopulation of 2.3% of patients with primary overt hypothyroidism who were unaware of their thyroid test results and/or any diagnosis of hypothyroidism³².

Heterogeneity in diagnostic approaches and lack of understanding of changes in thyroid hormone levels with increasing age have been cited as reasons for failure to diagnose hypothyroidism in India³³. These authors called for public health interventions at the national level to address these gaps in knowledge. Two surveys conducted in 2017 found low levels of understanding of the thyroid gland and its relationships with hypothyroidism among members of the public in India^34,35.

Clinical implications of unawareness of hypothyroid disease

Subclinical versus overt hypothyroidism

Current guidance for the management of subclinical hypothyroidism supports a trial of levothyroxine for patients aged <65 y with symptoms³⁶. Conversely, some older subjects or those without symptoms of hypothyroidism (presumably diagnosed opportunistically or via population screening for thyroid disease) are not considered to require thyroid hormone replacement^36,37. It could be argued that a diagnosis of subclinical hypothyroidism in the latter group would be moot, as it would not affect their management, while burdening them with an additional medical diagnosis. In addition, it will be important in future to identify and manage appropriately those subjects with modestly elevated TSH that is due to causes other than thyroid dysfunction, such as obesity and/or psychological stress, as described below^38,39.

A substantial proportion of people with subclinical hypothyroidism will progress to overt hypothyroidism (3–18% per year, according to one review⁴⁰). Identifying these subjects would facilitate investigation for risk factors for progression to overt disease, such as higher TSH or anti-thyroid peroxidize (TPO) antibodies³⁶.

Cardiovascular health

Subclinical hypothyroidism has been proposed as a modifiable risk factor for adverse cardiovascular outcomes, which may increase the urgency of diagnosing the condition^7,41–43. A recent report found that cardiovascular disease mediated about one-seventh of the excess mortality associated with subclinical hypothyroidism⁹. The relationship between subclinical hypothyroidism and outcome is probably complex however, and likely influenced by multiple factors including age, the severity and duration of exposure to abnormal thyroid hormone levels and, especially, obesity³⁸. Increased psychological stress has been proposed recently as an additional factor other that hypothyroidism that can result in modestly increased TSH levels, and which may also contribute to observations of increased cardiovascular risk associated with TSH in the high-normal range³⁹.

No randomized, controlled trial has yet demonstrated improved cardiovascular outcomes from levothyroxine replacement in patients with hypothyroidism. The Thyroid Hormone Replacement for Subclinical Hypothyroidism (TRUST) study did not demonstrate a reduction in overall cardiovascular events with levothyroxine vs. placebo in a population of older subjects with persistent subclinical hypothyroidism, although this was a secondary outcome and only 38 events were recorded⁴⁴. A pooled analysis of TRUST with another study in elderly people with subclinical hypothyroidism (IEMO80+: the Institute for Evidence-Based Medicine in Old Age 80-plus thyroid trial) reported similar results⁴⁵. It has to be taken into consideration that most patients recruited for TRUST had TSH values below 10mIU/L and had no clinical symptoms, thus representing the group of older patients with subclinical hypothyroidism that would not get levothyroxine treatment according to current guidelines. In addition, TSH levels tend to increase with age, so that there is a risk of misdiagnosing elderly people with hypothyroidism if age-appropriate reference intervals are not used^22,46. Randomized outcomes trial data from younger patients with subclinical hypothyroidism are required to help to guide the management of this population. Interestingly, observation data suggested a cardiovascular benefit associated with intervention with levothyroxine in younger, but not older, patients with ischemic heart diasease⁴⁷. It should be noted that it would be probably be unethical to conduct a randomized outcomes trial comparing LT4-based therapy with placebo in patients with overt hypothyroidism: there is a consensus that these patients require thyroid hormone replacement, for reasons that include preservation of cardiac function²⁴.

Although the magnitude of increased cardiovascular risk in subjects with high-normal TSH is small, large numbers of subjects fall into this diagnostic category⁹. Further research is required on the relationships between subclinical hypothyroidism and cardiovascular outcomes, and the extent to which this relationship is confounded by other factors, to support the need for interventions to protect the cardiovascular system in this population.

Other potentially adverse outcomes

Undiagnosed hypothyroidism has been identified as a risk factor for adverse outcomes in other settings. A case report described delayed recovery from general anesthesia, which was later attributed to the presence of undiagnosed hypothyroidism⁴⁸. Undiagnosed hypothyroidism has also been described as a risk factor for toxic reactions to digoxin treatment⁴⁹ (hypothyroidism delays the clearance of digoxin from the body and increases its concentration in the serum⁵⁰, and represents a caution to the use of this drug⁵¹).

Conclusions

The relatively consistent findings that about 5–7% of populations in the USA and Europe have undiagnosed hypothyroid disease implies a large number of individuals without a diagnosis who may be at risk of adverse clinical sequelae. For example, the American Thyroid Association estimates that some 20 million residents of the USA have some form of thyroid disease, and that as many as 60% – about 12 million people –are unaware of it⁵². It is important to diagnose thyroid conditions accurately, including patients with relatively mild subclinical hypothyroidism, so that they can be managed appropriately to optimize their long-term outcomes. This includes people with subclinical hypothyroidism who do not require immediate thyroid hormone replacement therapy, who can be followed for changes in the severity of the condition.

Improved education will be needed for the public as to the nature and symptoms of hypothyroidism. Expert societies already provide public awareness campaigns, such as the “Up To Here” campaign from the American Association of Clinical Endocrinologists (http://www.thyroidawareness.com/) and events such as “World Thyroid Day” and “International Thyroid Awareness Week” (https://thyroid-fed.org/events/). These initiatives are useful in reaching mainstream media. Patients with already diagnosed hypothyroidism are often young, motivated and avid users of social media, but campaigns aimed at increasing awareness of thyroid disease need to cut through to social media users who are not yet active consumers of healthcare information. It is also important that the messages provided are strongly evidence based, to counteract the many myths that circulate on the Internet and on social media, relating to thyroid disorders and other conditions.

Healthcare professionals need to diagnose or exclude hypothyroidism promptly when faced with a patient displaying the nonspecific symptoms associated with this condition. More research is needed to refine the diagnosis of hypothyroidism, with regard to the levels of TSH and other thyroid hormones, and a better understanding of how physicians may be able to incorporate symptoms, as well as biomarkers, into the differential diagnosis. This would aid the important clinical task of identifying more accurately the apparently substantial number of people with undiagnosed hypothyroidism, especially outside the USA and Europe and in regions of iodine depletion, or where the prevalence of hypothyroidism is high, such as India.

Supporting information

Search methods

This is a narrative review aimed at exploring available knowledge on the extent of unawareness and under diagnosis of hypothyroidism in adults. The review is based on a structured search of the PubMed database, limited to articles with full text in English. A series of broad initial searches combined” hypothyroid [ti]” with “diagnos* [ti] (295 hits), “(under diagnos* OR undiagnos*)” (120 hits), “aware” [ti] (126 hits), “unaware [ti] (11 hits). The results of these searches, and the reference lists of articles, were examined manually for articles of interest. Central, gestational and congenital hypothyroidism were considered to be beyond the scope of this review.

Transparency

Declaration of funding

Merck Healthcare KGaA funded editorial support (see below), Fast Track review, open access publication and color-printed figures for this article.

Declaration of financial/other relationships

UH is an employee of Merck healthcare KGaA. BS has declared no conflict of interest.

Acknowledgements

Dr Mike Gwilt (GT Communications) provided editorial assistance, funded by Merck Healthcare KGaA, Darmstadt.