An Online Survey of Hypothyroid Patients Demonstrates Prominent Dissatisfaction

Introduction

Hypothyroidism is a common endocrine problem that requires lifelong treatment with thyroid hormone (1). Currently, the standard treatment for hypothyroidism is daily administration of levothyroxine (LT4) at doses that normalize serum thyrotropin (TSH) (2,3). Even though thyroxine (T4) is intrinsically active in some settings (4), many tissues have deiodinases that activate T4 to triiodothyronine (T3), the biologically active thyroid hormone. The prevailing viewpoint is that a dose of LT4 that normalizes serum TSH resolves most/all hypothyroid symptoms (5).

However, a cross-sectional investigation of LT4-treated patients reported impaired psychological well-being compared to control individuals (6). The level of dissatisfaction of individuals on the general health questionnaire was 21% higher in LT4-treated patients compared to controls. Stronger differences (34%) were seen when a thyroid-specific questionnaire was used (6). Patients may complain of symptoms overlapping with hypothyroidism such as sluggishness, lethargy, sleepiness, memory problems, depression, cold intolerance, hoarseness, dry skin, body weight gain, and constipation (1). They may switch physicians multiple times and may use unconventional therapies, such as dietary supplements, nutraceuticals, and over-the-counter products. Symptoms consistent with hypothyroidism are nonspecific and cannot be used to identify hypothyroid patients among euthyroid controls (7,8). Hence, guidelines recommend withholding treatment with thyroid hormone when the diagnosis is not biochemically confirmed (2). Compared to healthy individuals who may report less than optimal quality of life (QoL) at rates of 25–35% (6), patients carrying diagnoses of chronic medical conditions report diminished QoL (9,10). Chronic conditions treated with daily medication further compromises patients' perception of the quality of their lives (11). Individuals treated with medications with predictable side effects further experience worse QoL (10). Thus, it is expected that patients with Hashimoto's thyroiditis too may complain of diminished QoL independent of thyroid function (12 –15). Among factors influencing the frequency with which a typical symptom such as fatigue will be reported in LT4-treated subjects are sex, employment status, and physical activity rather than thyroid function (16).

The existence of a subset of LT4-treated hypothyroid patients who have nonspecific symptoms who report poor QoL indicators has led to questions as to whether monotherapy with LT4 is adequate for all patients (3,17). Among National Health and Nutrition Examination Survey (NHANES) participants, LT4-treated patients with a normal serum TSH exhibit a 15–20% decrease in the ratio of circulating T3/T4 (18), and about 15% of these patients may not maintain normal serum T3 levels (19). This could result from differences in D2 regulation between the hypothalamus and peripheral tissues, as has been shown in a hypothyroid rat model (20). In fact, replacement with LT4 might not fully restore all aspects of euthyroidism, as seen in females taking LT4 who have lower energy expenditure compared to euthyroid women with similar age and body mass index (BMI) (21). In addition, the NHANES data cited above (18) indicate that individuals taking LT4, with TSH values within the normal reference range, have a higher BMI, despite reporting lower calorie intake corrected by body weight, in the setting of a reported lower physical activity levels. Additionally, the LT4-treated subjects were significantly more often taking statins, beta-blockers, and antidepressants compared to euthyroid participants matched by age, sex, and race/ethnicity (18). Following additional statistical analysis, it was demonstrated that LT4-treated subjects reported an increased frequency of episodes of memory problems/confusion and were less likely to report being in excellent/good health (22). An attempt to enhance effectiveness of thyroid hormone replacement with higher doses of LT4 may explain at least in part why some patients may exhibit suppressed TSH (23).

Fueled by a hypothesis that subnormal T3 levels contribute to residual symptoms, a total of 14 double-blind placebo-controlled trials of combination therapy have been conducted. Based on these trials, it is not clear that combination therapy is superior to monotherapy in managing hypothyroidism. These trials provided heterogeneous results with respect to health-related QoL and mood and neurocognitive functioning, although there was a patient preference for combination therapy in some of the trials (6,24 –36). A single randomized trial of desiccated thyroid extract (DTE) therapy compared to LT4 failed to identify differences in neurocognitive testing or in symptoms (primary outcome) but showed weight loss and treatment preference for DTE (37). Currently the American Thyroid Association (ATA) guidelines do not recommend the routine use of combination therapy (2).

The objective of the current analysis was to probe the perceptions, comorbidities, and treatment selections of survey respondents who expressed a low overall degree of satisfaction with their therapy.

Methods

Results

A total of 12,146 respondents completed the survey, and all subsequent analyses were performed based on the self-reported responses to questions about medical history and hypothyroidism treatment. Respondents who were not taking medication for hypothyroidism (n = 53) were excluded. Of the remaining 12,093 individuals (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/thy), 485 were excluded because they were taking a medication for hypothyroidism other than LT4, LT4 + LT3, or DTE. An additional 442 were excluded due to survey completion from the same IP address and concern that the data represented duplicate surveys. As a result, the total sample of respondents that was further analyzed comprised 11,166 individuals (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Flow diagram of survey respondent eligibility.

The female:male ratio of respondents was approximately 21:1, and age was relatively evenly distributed across the four age categories (Table 1). The most prevalent cause of hypothyroidism was Hashimoto's/autoimmune disease (43%). However, 34% of respondents identified another cause other than Hashimoto's/RAI/surgery, or were unsure of hypothyroidism etiology (Table 1). Only 7% of individuals were treated for hypothyroidism for less than one year; the majority (63%) had been on treatment for more than six years. One third of patients stated that their current stress level could be contributing to hypothyroid-related symptoms; another third reported co-existing medical conditions, with the most common one being depression (27%). Only 6% of respondents self-reported depression without any other comorbidities (Table 1).

Within the total sample, 6949 individuals were taking LT4 monotherapy, 978 reported taking combination LT4 + LT3, and 3239 received DTE (Table 1). When considering perceptions regarding their treatment for hypothyroidism, on a scale from 1 to 10, with 10 being the highest, the median response indicating treatment satisfaction was 5 (IQR = 3–8; Table 1). Among those who were frustrated with their hypothyroidism treatment, the relevant areas identified as causing dissatisfaction were weight management (69%), fatigue or energy level (77%), mood (45%), and memory or other problems thinking (58%). The median response describing satisfaction with the patient's current physician was 6 (IQR = 3–6), and assessment of the doctor/physician knowledge regarding hypothyroidism treatment was 5 (IQR = 3–8); 54% of the sample reported changing physicians more than twice because of dissatisfaction with treatment. Almost all respondents believed that there was a strong need for new treatments for hypothyroidism (median 10 [IQR = 8–10]) and perceived a significant influence of hypothyroidism on life (median 10 [IQR = 8–10]; Table 1).

Next, multiple analyses were utilized to compare the responses within the total sample according to their specific treatment (Table 2). When examining the three treatment subgroups, the distribution of sex, age, etiology, and treatment duration were significantly different. Individuals treated with DTE had the highest median satisfaction with treatment (7 [IQR = 4–8]) compared to those taking LT4 (5 [IQR = 3–7]) or LT4 + LT3 (5 ([IQR = 3–7]). In particular, as shown in the graphic representations, the distribution of responses was markedly different between those taking DTE versus LT4 (Table 2). Individuals on DTE predominately responded positively with an upward trend, such that responses were more frequent at the positive end of the scale. Conversely, individuals on LT4 were more likely to respond negatively, exhibiting a distribution with a downward trend, such that responses were more frequent at the negative end of the scale (Table 2). This varied distribution between groups can also be described by examining the number of patients who ranked satisfaction with treatment as not satisfied (ranked 1 or 2) or completely satisfied (ranked 9 or 10). Approximately 20% of respondents taking LT4 and LT4 + LT3 were not satisfied with treatment, while 14% of DTE users were not satisfied. In comparison, 22% of DTE users were completely satisfied compared to 10% of LT4 and LT4 + LT3 respondents (Supplementary Fig. S1). Individuals taking DTE were less likely to report problems with weight management, fatigue/energy level, mood, or memory when compared to those taking LT4 or LT4 + LT3 (Table 2). Individuals taking DTE had a higher median satisfaction with their current physician (7 [IQR = 4–9]) compared to those taking LT4 (5 [IQR = 3–8]) or LT4 + LT3 (6 [IQR = 3–8]). Perceived physician knowledge was slightly higher in the DTE subgroup compared to the LT4 subgroup (Supplementary Figs. S2 and S3). Of note, 29% and 21% of individuals taking DTE and LT4 + LT3 changed doctors five or more times because they were not satisfied with their treatment compared to only 7% of respondents taking LT4. Those taking DTE or LT4 + LT3 were more likely to have tried alternative treatment forms not prescribed by their doctor and thought their lives had been more affected by hypothyroidism (10 [IQR = 9–10]), although this latter parameter was very high in all three subgroups (Table 2).

The univariate analysis testing the association between demographic data and the characteristics of the disease revealed that the following parameters were associated with a higher likelihood of treatment dissatisfaction (odds ratio [OR] >1.0): (i) female; (ii) changing physician twice or more; (iii) complaints about weight management, fatigue or low energy levels, mood, and memory; (iv) tried alternative treatment for hypothyroidism; and (v) depression (Table 3). In contrast, (i) being aged >60 years, (ii) having hypothyroidism due to RAI or surgical thyroidectomy, and (iii) being on LT4 + LT3 or DTE therapy were associated with a lower likelihood of treatment dissatisfaction (OR <1.0; Table 3). These associations remained statistically significant, even when testing was done through multivariate analysis (Table 4). Independently of type of treatment, concerns over weight management, fatigue, mood, and memory were associated with dissatisfaction (OR >1; Supplementary Table S2).

Discussion

The present study reports the results of a large-scale assessment of patients' perceptions about hypothyroidism. It should be emphasized that the study is based on responses provided by a self-selected sample that is unlikely to represent the >10 million Americans being treated for hypothyroidism. For example, we know that women comprise approximately 80% of all hypothyroid cases (38), whereas here a 21:1 ratio was observed between female and male respondents (Table 1). Also, based on the authors' experience, <10% of the hypothyroid patients are on DTE, and yet here it was observed that 29% of the respondents were taking DTE (Table 1). Thus, a number of elements in the cohort are acknowledged that indicate this is not a bias-free population. However, the self-reported issues with weight management, fatigue/energy levels, mood, and memory observed in this study corroborate findings reported previously (6), as well as the observation of increased BMI and issues with memory and confusion observed in LT4-treated hypothyroid individuals (18,22).

The results of this survey suggest that dissatisfaction with hypothyroidism treatment and treating physicians are important problems for patients. Furthermore, a strong need for development of new treatments for hypothyroidism was identified. These are noteworthy findings, as among physicians, treatment of hypothyroidism is considered to be straightforward. The median reported satisfaction with treatment in the entire group of only 5 on a scale of 1–10 is remarkable and concerning, even if this only reflects the situation in a small portion of patients. In light of the high prevalence of hypothyroidism, the findings suggest a significant burden of unsuccessfully resolved symptoms across the population, particularly in women. At face value, these results indicate that although physicians believe that hypothyroidism is an eminently treatable condition, a portion of such patients believe their lives have been greatly affected by the disease and are unhappy with their treatment and physicians. Almost universally, they believe there is a need for the development of new treatment forms. It is also remarkable that there is a clear preference for DTE in the whole group, as well as when the group was broken down into multiple subgroups. Of course, the study is limited by the potential intrinsic sample bias. However, the suggestion that something “real” is being captured is bolstered by the finding that the survey did not demonstrate a clear positive patient response to synthetic combination therapy with LT4 + LT3.

The focus of prior research into combination therapy has been on synthetic LT4 + LT3 rather than DTE. The 14 trials of synthetic LT4 + LT3 that have been completed thus far show some patient preference for combination therapy, but have failed to show obvious superiority of LT4 + LT3 (6,24 –30,32 –36). Thyroid-related symptoms were generally not improved with combination therapy, other than when TSH suppression was achieved. Parameters such as QoL, mood, and neurocognitive performance were only improved in a minority of studies. These studies each have limitations (e.g., once daily dosing, short duration study, small study size, disparate TSH values between study groups) that have been previously reviewed extensively (2,39). Failure to demonstrate superiority of LT4 + LT3 could be due to any combination of these shortcomings in study design or the drug formulation. However, it is also possible that synthetic combination therapy is not superior to LT4. The one double-blind, randomized, placebo-controlled trial of DTE versus LT4 also failed to show that DTE resulted in improvement in a number of neuropsychological measures or symptoms (37). Two secondary outcomes—preference for DTE and modest weight loss—were associated with DTE use. However, participants only took each therapy for 16 weeks in the study, and long-term data were not reported. A preference for LT4 + LT3 has also been shown to be associated with the weight loss achieved during therapy, although TSH suppression was a confounding factor (24). In an attempt to query patients, a Danish Internet-based questionnaire surveyed 293 individuals on combination therapy, revealing that 84% of patients who had residual symptoms while on monotherapy perceived improved symptoms after switching to combination therapy, and 81% stated a clear preference for continuing combination therapy (40). Few physicians are willing to prescribe combination therapy to manage residual symptoms (41), and minimal information is available to describe prescription patterns of DTE.

If the randomized, placebo-controlled clinical trials have not shown benefits of either LT4 + LT3 or DTE, but patients surveys (40) and online patient forum opinions suggest that combination therapy is preferred, this reflects either uncontrolled bias or lack of identification of the appropriate patient group for clinical trial enrollment. Prospective clinical trials of combination therapy have not been conducted that have specifically recruited dissatisfied patients or patients with the lowest circulating T3 levels. No trials have incorporated assessment of deiodinase or thyroid hormone transporter polymorphisms into their primary design. Retrospective data suggest that patient preference may be linked to a patient's complement of thyroid hormone metabolism-associated polymorphisms (42,43).

If DTE provides more satisfactory therapy for patients with hypothyroidism, it is possible that this is due to (i) patient preference for higher treatment doses, (ii) patients being rendered T3 thyrotoxic, (iii) the presence of some other orally active substance other than T4 and T3 within the DTE, (iv) a confounding factor such as use of other complementary or alternative medicine in users of DTE, or (v) an as yet unidentified aspect of thyroid physiology. It is important to keep in mind though that DTE, like LT4, does not restore normal thyroid hormone homeostasis. Circulating levels of T3 are increased during DTE therapy and may transiently exceed the upper limits of normal, while the average blood levels of T4 are below the lower limit of normal. High levels of T3 are known to enhance mood in studies of patients with depression, and it is possible that patient preference for DTE reflects a positive effect of supra-physiologic T3 levels on mood. At the same time, it is unknown whether transient supra-physiologic T3 levels are safe or whether they could promote arrhythmias, especially in older or susceptible patients. With respect to patients potentially preferring higher doses, it has recently been shown in a randomized blinded trial that patients preferred the LT4 dose that they believed was the highest, even if they had not correctly identified this is the highest dose (44).

A major strength of this study is the large sample size. However, limitations of the study include sampling and/or recall bias, subjectivity, a lack of an external control (e.g., patients treated for other endocrine disorders or other chronic medical conditions), and the use of a non-validated survey instrument. With respect to the first limitation, the increased likelihood that patients with significant dissatisfaction with their therapy for hypothyroidism were more likely to have been motivated to complete the survey than those who felt unaffected by their hypothyroidism, or even those who felt happy with their treatment, is acknowledged. To highlight this, 50% of the survey respondents had changed their physician twice or more. In addition, two thirds of respondents believed that stress or physical comorbidities other than hypothyroidism might account for at least part of their symptoms. It is anticipated, although not verified, that this questionnaire attracted dissatisfied patients or was preferentially publicized among groups of dissatisfied patients. It is therefore anticipate that one end, or possibly both ends, of the spectrum of opinion about treatment have been captured. If it is assumed that approximately 15% of treated patients feel worse than individuals without thyroid disease, and if a subset of these patients has been captured, the results, despite the inherent bias, nevertheless indicate a significant unmet need among patients. The study also seems to have captured predominately females. It is known that hypothyroidism affects women and men at a ratio of 9:1, and yet the respondents exhibited a ratio of 21:1.

With respect to the third and fourth limitations, because the diagnosis of hypothyroidism is self-reported, it is not possible to say for sure whether the respondents do not include a significant number of individuals who are taking thyroid hormone because of a misdiagnosis of hypothyroidism or for a condition other than hypothyroidism, for example fibromyalgia. Additionally, hypothyroid patients may also mistakenly attribute unrelated symptoms or decreased QoL to their thyroid condition. Once a patient has been diagnosed with a chronic condition such as hypothyroidism, there is a natural tendency for a patient to associate their spectrum of symptoms with this condition. The attribution of these symptoms may be mistaken. If there is mis-attribution of symptoms, then these symptoms would not be expected to resolve with adjustment of therapy for hypothyroidism. It is well known that patients with hypothyroidism have a greater disease burden than the general population (6,45), and that hypothyroid patients treated to achieve a normal TSH may remain symptomatic (46,47). However, not only does manipulation of thyroid status in these individuals fail to resolve symptoms (35,44), but also treatment of euthyroid individuals with hypothyroid-like symptoms does not resolve symptoms (48). An attempt was made to mitigate these particular limitations by requesting that respondents report co-existent medical conditions and by examining subgroups who did not have these conditions. In general, the findings remained generally unaltered in these subgroup analyses. However, it is not known if other medical conditions have been fully captured that might be the major source of some of the symptoms reported.

In conclusion, it is clear that a subset of patients with hypothyroidism are not satisfied with their current therapy or their physicians. Future surveys should aim at sampling representative cohorts of hypothyroid individuals to describe patient experience and satisfaction with treatment globally. The overall limitation of the studies performed so far highlights the need for high-quality research to study treatments for hypothyroidism. Such treatments may include hormonal therapies, supportive care interventions, life-style modification interventions (e.g., exercise, diet), or complementary/alternative treatments. Definitive trials need to be adequately statistically powered to detect clinically significant changes in important patient outcomes, attempt to provide steady levels of T3, and specifically target individuals who are symptomatic. Failure to conduct well-designed studies to advance understanding in this area promotes reliance on anecdotal case reports/series, self-report survey studies (such as this one), and observational registry data. The QoL issues raised by this survey need to be considered when making medical decisions to render patients permanently hypothyroid, and these issues should be discussed with patients. More research is also needed into weighing the risks and benefits of strategies to preserve thyroid function in situations where this may be a reasonable option (e.g., active surveillance of papillary microcarcinoma, or use of hemithyroidectomy instead of total thyroidectomy in low-risk differentiated thyroid cancer). In the absence of a better understanding of hypothyroidism treatment, patients will continue to experience unresolved symptoms and be exposed to the risks and expenses of treatments with unproven benefits and possible harm (49).