The original goal of this project was to see if symptoms could distinguish between people with thyroid dysfunction and those who are euthyroid. We did not find this to be the case in this population. No symptom, nor group of symptoms, predicted hypothyroidism. Symptoms did not discriminate between disease states because both euthyroid and hypothyroid individuals reported similar symptom prevalence. Interestingly, the euthyroid people who presented to this thyroid screening health fair reported a high proportion of classic thyroid symptoms, higher than a group of euthyroid individuals in a previously published case-control study.
The secondary goal of the project (to determine the yield of thyroid screening in this population as compared with previous studies) found the yield of thyroid disease on testing this group of health fair attendees higher than previously reported in the literature . This screening program identified 97 people (12.2%) with newly diagnosed thyroid dysfunction, 34% of which had a TSH greater than 10 μIU/ml. The 9.7% newly identified with hypothyroidism greatly exceeds the prevalence rates of hypothyroidism quoted in the literature (one to two percent of the general population) . Our rates are higher even when the prevalence of subclinical hypothyroid is included in the comparison studies . Interestingly, the rates of an elevated TSH found in our study are quite comparable to those reported in another health fair population .
There has been much discussion regarding traditional symptoms of thyroid disease and the utility of symptoms in detecting hypothyroidism. Such symptoms are not highly sensitive, though multiple investigators have tried to quantify symptoms to aid in diagnosis [13, 21–23]. As well, symptoms may be too nonspecific to be helpful clinically, particularly in individuals with multiple comorbid conditions. Certainly individuals with co-morbid conditions manifest symptoms that could be attributed to thyroid disease. Information regarding comorbidities was not collected through the health fair, a limitation of our study.
The prevalence that symptoms were reported by all participants in this health fair study was considerably higher than the prevalence of symptoms previously reported in a case control study using the same symptoms questionnaire . Every questionnaire symptom was reported more often by both hypothyroid and euthyroid participants in this screening study, than by the euthyroid controls in the previously published study. Several explanations are possible. Foremost, the thyroid health fair was advertised with promotional materials that highlighted the symptoms associated with thyroid disease. This certainly could have attracted people with more typical thyroid symptoms. Health fairs vary in focus [3, 4, 9, 24, 25], and as Lefebvre and colleagues showed, different screening offerings do attract different populations . It is very likely that a health fair that advertises thyroid screening would attract people who are concerned about thyroid disease. Their concern may be related to symptoms attributable to thyroid disease, or to other factors such as family history of thyroid illness. People with a family history may be more aware of symptoms classically associated with thyroid dysfunction, and may want to participate in a thyroid screening. So, the higher prevalence of thyroid symptoms and of thyroid dysfunction found in our study may be explained by an increased awareness of personal symptoms and the high proportion of affected family members (42.4%). Symptom awareness was also likely enhanced by the educational activities during the thyroid health fair week. The increased prevalence of thyroid dysfunction found in our study may also be explained in part by the fact that health fairs typically attract individuals who are older and who are women [5–9]. This may have contributed to our findings since thyroid dysfunction, and hypothyroidism in particular, is more common in older people and is nearly ten times as common in women as in men . The populations which are at higher risk for thyroid disease are over-represented in our health fair study (and at health fairs in general) as compared with the general population.
It is interesting that among euthyroid individuals, some symptoms were still reported significantly more often by euthyroid individuals taking thyroid medication than individuals that were naturally euthyroid. Despite correcting biochemical hypothyroidism with medication, these individuals still reported a higher prevalence of some classic hypothyroid symptoms than biochemically euthyroid individuals who had not had thyroid disease. Whether this reflects heightened awareness of classic symptoms of hypothyroidism, or if it suggests that medication does not completely reverse hypothyroid symptoms is unknown.
We also observed that almost half of the people identified as having abnormal thyroid function through this screening program were people taking thyroid medication but who were not euthyroid despite being on treatment. This finding is consistent with the literature , reinforcing the need for closer monitoring of patients on thyroid hormone replacement.
Our study was limited by free T4 levels not being available through this screening. We are therefore unable to comment on the amount of subclinical hypothyroidism as defined by an elevated ultrasensitive TSH but normal free T4 level. As well, there has been discussion that the target euthyroid TSH may actually be less than the upper limit of the TSH assay. While multiple studies have looked at clinical manifestations associated with these milder forms of thyroid dysfunction, it is certainly beyond the scope of this study. When viewed conservatively, that is at a cutoff commonly used to represent overt disease, the nearly six percent of the individuals we screened who had a TSH level elevated greater than 10 μIU/ml warrants consideration. Using a lower cutoff of TSH to define hypothyroidism would only enhance the yield of testing in our health fair population. Regarding symptoms, it is unknown but unlikely to affect results since there were no statistically significant differences in symptom reporting between euthyroid and hypothyroid individuals. We are also limited by having data from one point in time, as is the nature of health fairs. This does not allow us to know if any TSH results reflected transient abnormalities. However, it is unlikely that such deviations would favor abnormal or normal thyroid function in particular.
Thus, testing people who wish to be evaluated for thyroid disease can increase diagnostic yield. In our study, testing a self-selected population increased identification of previously unknown thyroid dysfunction from the less than 2% quoted for the general population to 12% in this health fair population. The observation that more disease is identified through a disease-specific health fair than is reported in the general population, we call the “health fair effect”. This may reflect education about traditional thyroid symptoms, impact of affected family members, or the characteristics of the people themselves having more risk factors such as age and female gender. While testing symptomatic people may increase yield, it did not allow discrimination between disease states in this particular population. Traditional thyroid symptoms were highly prevalent in all people who attended this health fair, so did not aid in identification of disease. It is possible that symptoms would be more discriminatory in a population that had not been educated on thyroid dysfunction, or in a setting outside of a health fair where people are motivated often by risk factors for the disease being screened. The application of symptoms to direct thyroid testing may be better suited to a clinical setting rather than a health fair, a possible direction for future research.
Thyroid screening does identify people who may otherwise go undiagnosed and thus untreated, and who may benefit from treatment because of the adverse effects of thyroid disease on multiple organ systems. But thyroid testing needs to be done in a setting where the chance of identifying disease is high enough to be beneficial. Testing a population with a likelihood of thyroid disease that is greater than the general population, is more desirable. The effect of this health fair was to draw a population more likely to have thyroid disease, for several possible reasons, and thus increase the yield of testing for thyroid disease. Whatever drew people to have thyroid function tested at this Thyroid Awareness health fair, resulted in an enriched prevalence of thyroid disease. This health fair effect may be explained by people attending the health fair because of an increased concern for thyroid disease and desire to be tested. Such concerns may be multifactorial, perhaps reflecting an increased awareness of disease symptoms because of family history of thyroid disease, and/or the information presented by the media during Thyroid Awareness Week. Attracting people with a high proportion of suspect symptoms may contribute to the health fair effect in this study. The health fair also drew a greater proportion of women participants, and participants who are older than the general population. This contributed to the health fair effect by attracting a demographic known to have a higher likelihood of thyroid disease. Other unmeasured factors may also exist. Thus, diagnosing previously unknown thyroid dysfunction through testing of people who present to disease-specific health screenings may be both effective and appropriate. Trying to discriminate between individuals with and without thyroid dysfunction solely based on symptoms may be more appropriate in other settings, such as primary care clinics. Further studies are needed to look at the cost-effectiveness of thyroid screening in health fair populations, and populations that resemble health fair attendees, who have a greater likelihood of disease.