- The patient’s QoL has not been an issue for physicians to study for most of medical history.
- Shameful neglect of QoL studies of thyroid patients that are not hyper or hypothyroid but just goiters or nodules.
- Asymptomatic patients may suffer a reduction in perceived health status due to distress related to physical appearance and awareness of disease.
- 368 patients (hypothyroid, 81; hyperthyroid, 45 (for both states including overt and subclinical states); Hashimoto thyroiditis, 51; euthyroid goiter, 191) were studied.
- HRQL was impaired also in the absence of altered hormone levels.
- HRQL was impaired also in the absence of altered hormone levels.
- Mood/behavior disturbances were present in a large proportion of patients and were significantly associated with poor HRQL.
- HRQL was significantly reduced in patients with thyroid diseases referred to an endocrinology.
- Perceived health status may be considered as an additional outcome of management and therapy of thyroid disorders.
- (HRQL), subjectively perceived by the patient, is becoming a key component in the evaluation of any therapeutic intervention.
- Patients may be even more concerned about quality and disability than about longevity, mainly in chronic diseases where survival for a long time, is not at risk and the goal of interventions is to maintain patients symptom-free and living in the community.
- The physician must take QoL issues seriously when treating long standing goiter,nodules or cancer.
- If you have Qol issues unresolved during your visits for thyroid related conditions consider changing to a thyroidologist.
- 310393-8860 or email to firstname.lastname@example.org.
- Ask for Alicia Dr.G.
Health-related quality of life in patients with thyroid disorders
Quality of Life Research 13: 45–54, 2004 G.P. Bianchi et al
Factors significantly associated with poor HRQL in relation to clinical status in subjects with thyroid disorders (odds ratio
and 95% confidence interval)
Euthyroid N = 288 Hypothyroid N = 49 Hyperthyroid N = 31
Physical functioning Paresthesias (2.8, 1.6–5.0)
Sleep (2.0, 1.1–3.6)
Fatigue (9.0, 2.0–40.8) –
Role limitation-physical Fatigue (3.0, 1.7–5.3)
Bowel habits (1.9, 1.1–3.2)
Fatigue (8.3, 1.8–38.1) Pruritus (6.5, 1.3–33.0)
Role limitation-emotional Behavior (2.3, 1.4–3.9)
Fatigue (2.2, 1.2–3.9)
Behavior (3.5, 1.0–12.1) Pruritus (6.9, 1.2–40.4)
Bodily pain Behavior (1.9, 1.1–3.3) – Cramps (17.2, 1.5–210.2)
Vitality Fatigue (3.0, 1.7–5.5)
Behavior (2.5, 1.4–4.4)
Fatigue (16.3, 2.3–112.7)
Behavior (7.7, 1.4–42.9)
Behavior (6.5, 1.3–33.0)
Mental health Behavior (5.6, 2.7–11.6)
Fatigue (1.9, 1.0–3.6)
Behavior (14.2, 1.7–121.8) Cramps (9.5, 1.4–64.4)
General health Behavior (4.6, 2.5–8.5)
Paresthesias (1.8, 1.1–3.1)
Behavior (11.6, 1.4–100.2) Sleep (10.5, 1.7–63.9)
Social functioning Behavior (3.2, 1.8–5.7)
Paresthesias (1.8, 1.0–3.4)
Behavior (30.4, 3.5–263.8) –
Energy Sleep (10.0, 4.4–23.0) – Cramps (7.5, 1.1–49.3)
Emotional reactions Sleep (3.2, 1.5–7.1)
Behavior (3.0, 1.3–6.8)
Paresthesias (12.9, 1.4–118.6) Cramps (8.2, 1.0–66.2)
Sleep Behavior (4.9, 2.1–11.5)
Fatigue (3.0, 1.5–6.0)
Sleep (2.5, 1.2–5.1)
Pruritus (4.8, 1.1–22.3) Sleep (10.0, 1.0–97.1)
Social isolation Behavior (2.8, 1.5–5.1)
Fatigue (2.2, 1.2–4.0)
Behavior (9.0, 1.0–78.2) Behavior (8.0, 1.3–48.2)
Mobility Behavior (2.5, 1.3–4.8)
Fatigue (2.4, 1.3–4.4)
– Sleep (8.0, 1.3–48.2)
Pain Behavior (3.7, 1.9–7.2)
Sleep (2.2, 1.2–4.0)
Fatigue (2.1, 1.1–3.9)
– Behavior (6.1, 1.0–36.9)
Abstract Limited reports are available on quality of life (HRQL) in thyroid diseases, and no data are available in euthyroid disorders, such as goiter and Hashimoto thyroiditis. Also, asymptomatic patients may suffer a reduction in perceived health status due to distress related to physical appearance and awareness of disease. We measured HRQL by means of Medical Outcome Study Short Form-36 (SF-36) and Nottingham Health Profile (NHP) questionnaires in 368 patients (hypothyroid, 81; hyperthyroid, 45 (for both states including overt and subclinical states); Hashimoto thyroiditis, 51; euthyroid goiter, 191). The final scores of the domains were compared with age- and sex-adjusted Italian normative values, by computing the effect size. All domains of SF-36, except bodily pain, were reduced in thyroid disease; this was mainly the case of role limitation (both physical and emotional), general health and social functioning. The domains of NHP were less severely affected. HRQL was impaired also in the absence of altered hormone levels. Mood/behavior disturbances were present in a large proportion of patients and were significantly associated with poor HRQL. HRQL was significantly reduced in patients with thyroid diseases referred to a secondary level endocrinology unit. Perceived health status may be considered as an additional outcome of management and therapy of thyroid disorders. Key words: Goiter, Mental Health, Physical Health, Quality of life, Thyroid Introduction Health-related quality of life (HRQL), subjectively perceived by the patient, is becoming a key component in the evaluation of any therapeutic intervention. In some instances, patients may be even more concerned about quality and disability than about longevity , mainly in chronic diseases where survival for a long time, is not at risk and the goal of interventions is to maintain patients symptom-free and living in the community. Thyroid disorders are an example of such diseases. Their prevalence is high in the community [2, 3], and also, in Italy, a high prevalence has been reported . Only a few patients are overtly symptomatic; in euthyroid diseases, such as thyroiditis or goiter, many patients may remain asymptomatic for years or decades. In these subjects, therapeutic decisions are usually based on hormone levels, and patients’ opinions on the perceived health status are scarcely considered by clinicians. Only a few data are available on HRQL in thyroid diseases. Most of them were collected from patients with clinically overt hypothyroid [5, 6] and hyperthyroid diseases [7–10], or in Graves’ ophthalmopathy [7, 10, 11], where altered HRQL is more likely the effect of ocular pathology than of thyroid hormone excess. Both general [7, 10] and disease-specific  questionnaires have been used. Quality of Life Research 13: 45–54, 2004. 2004 Kluwer Academic Publishers. Printed in the Netherlands. 45 In spite of the lack of significant symptoms specific to altered thyroid function, the burden of disease on patients’ well-being might also be relevant in euthyroid patients. In euthyroid goiter, patients may suffer the discomfort of neck enlargement or upper chest occupation, contributing to the poor perceived health status and adding to the distress generated by the physical appearance or simply by a diagnosis of thyroid disease. We assessed HQRL in benign thyroid diseases by means of the Italian version of the Medical Outcome Study Short Form-36 (SF-36)  and the Nottingham Health Profile (NHP) . These generic indices were used in an attempt to verify the possible burden of asymptomatic thyroid disorders on HRQL. Clinical and laboratory data were also collected in order to determine the most relevant factors associated with loss of physical and emotional health in a disease having a large prevalence in the Italian population. Materials and methods Patients The study on HRQL in thyroid diseases started in March 1997. During a period of 8 months, we enrolled 368 consecutive, community-living, adult patients with newly diagnosed, untreated thyroid disease (323 F, 45 M; aged 18–85 years), independently of socio-demographic characteristics, educational and clinical status. Eighty-one patients were diagnosed as having hypothyroidism, 45 as having hyperthyroidism, 51 as having euthyroid Hashimoto thyroiditis and 191 as having euthyroid goiter. Patients had been seen in our secondary-level referral Unit either for clinical symptoms or because of occasional finding of altered hormone levels. The final diagnosis was always based on clinical data and an extensive battery of laboratory tests, with or without histology. Their sociodemographic, clinical and laboratory data are reported in Table 1. Age distribution was similar in the different groups: Only 19 out of 368 patients (5.2%) were aged 75 years or more; the prevalence of old people did not differ within groups. Among clinical data, we considered symptoms usually present in patients with thyroid diseases, namely pruritus, muscle cramps, abnormal bowel habits, associated diseases, as well as sleep and behaviour disturbances. Sleep disturbances included late sleep and early awakening; behaviour disturbances included anxiety and depression, as well as panic attacks. Thyroid function was extremely variable, ranging from severe disease to subclinical hyper- or hypo-thyroidism. In particular, 45 patients with hypothyroidism were considered to have subclinical disease on the basis of detecting elevated thyroid-stimulating hormone (TSH) and normal free thyroxine levels, whereas 26 had overt disease (elevated TSH and lower-than-normal free thyroxine). Similarly, among hyperthyroid patients, 27 were overtly hyperthyroid (elevated free thyroxine and/or triidothyronine), and 18 had normal free thyroxine with suppressed TSH (subclinical hyperthyroidism). Among patients with thyroiditis, 23 had laboratory evidence of having hypothyroidism, four had increased free thyroid hormone levels. All goiter patients had normal thyroid hormone concentrations. Associated diseases were diagnosed based either on previous medical records or on specific drug treatment. Diabetes (17 cases), depression (8 cases) and peptic ulcer (16 cases) were the most commonly associated diseases. A specific structured interview was used to detect mood/behaviour disturbances, namely anxiety, depression, listlessness, and panic attacks, as well as alterations in sleep and bowel habits, which are frequently associated with thyroid diseases. Over 97% of the subjects agreed to take part in the study and to fill the questionnaires, after signing an informed consent; the study was approved by the ethical committee. Methods The SF-36 and NHP questionnaires were presented to patients by nurses and completed before the clinical examination. They were self-administered, but the personnel were trained to check the completeness of data and to interview patients on missing data, without forcing answers from them. Therefore, data sets were almost complete, and the individual domains of HRQL were computable in 353 cases or more. The two questionnaires were administered in random sequence, and medical 46 personnel carried out clinical data recording after the visits. Data analysis First, data obtained in thyroid disorders were used to compute the scores of individual domains of the two questionnaires, as well as composite indices (physical component summary (PCS) and mental component summary (MCS) of SF-36 and NHP total score). Secondly, data were analysed by means of Cronbach’s coefficient a , and estimates for subscales gave a values ranging from 0.73 to 0.91 for SF-36 and from 0.75 to 0.85 for NHP, always exceeding the conventional threshold of 0.70 for comparison between groups . Following this, values measured in single diseases were compared with control values obtained in two large Italian population studies. Control databases for SF-36 and NHP included 2031 and 2248 subjects, respectively [12, 13], randomly selected via registry offices in two areas of Northern Italy, independently of their present health status. The age and gender distributions had been balanced to make these populations representative of the whole Italian population. Therefore, these values represent the average HRQL of the general Italian population. Data were averaged according to sex and age in the following age-ranges: O24, 25–34, 35–44, 45–54, 55–64, 65–74, P75 years. Statistical analysis The value of any individual domain of each patient was compared to the age- and sex-matched control group using the Z-score (difference between patient value and control mean, divided by control standard deviation). Z-scores were grouped according to sex, age, pathology, clinical status, complications of disease and symptoms. The mean and 95% confidence interval for each patient group and for each domain was calculated. A mean Z-score (effect size) <0.20 was considered clinically non-significant, a Z-score in the range 0.2–0.5 was classified as small, that between 0.5 and 0.8 as moderate, and that >0.8 as large [16, 17] (Table 2). To detect factors associated with poor HRQL, a logistic regression analysis was performed using dichotomized Z-scores of individual domains as dependent variables. The cut-off value was set at )0.8 for SF-36 and þ0:8 for NHP, but sensitivity analyses, using cut-off of 0.5 and 1.0, were also performed, and the results were qualitatively confirmed (not reported in detail). Independent variables were mood/behaviour disturbances, Table 1. Socio-economic, clinical and laboratory parameters of thyroid patients Hypothyroidism N = 81 Hyperthyroidism N = 45 Goiter N = 191 Thyroiditis N = 51 Gender (M/F) 3/78 11/34 28/163 3/48 Age (years) 54 [18–83] 60 [22–84] 55 [18–85] 45 [18–85] Marital status (%) (Single/Married/Widowed or Divorced) 17/73/10 20/76/4 19/69/12 27/67/6 Education (%) Primary/Secondary/Commercial or Vocational/Degree 26/43/19/12 36/23/36/5 43/26/26/5 22/32/41/5 Pruritus (%) 32 (22–43) 36 (22–51) 22 (16–28) 24 (13–37) Cramps (%) 40 (29–51) 38 (24–53) 35 (28–42) 27 (16–42) Bowel disturbances (%) 41 (30–52) 33 (20–49) 37 (30–44) 43 (29–58) Sleep disturbances (%) 63 (46–71) 44 (30–60) 57 (50–64) 41 (28–56) Behavior disturbances (%) 65 (52–73) 47 (32–62) 58 (51–64) 63 (48–76) fT3 (pg/ml) 2.3 (0.2–3.1) 8.7 (3.9–21.1) 2.9 (2.1–4.0) 2.9 (2.1–4.1) fT4 (pg/ml) 8.5 (2.2–11.6) 34.5 (10.1–126.0) 11.1 (8.0–17.9) 11.2 (8.0–19.0) TSH (U/l) 8.0 [5.5–62.7] 0.04 [0.0–0.30] 1.36 [0.40–4.22] 2.15 [0.40–4.40] Anti-TPO (U/l) 1416 [50–65,541] 725 [50–20,458] 85 [38–250] 1181 [50–98,000] Anti-TG (U/l) 100 [17–2400] 67 [4–2928] 83 [13–87,550] 100 [1–87,550] HTG (ng/ml) 38.5 [1.4–95.0] 22.2 [0.7–62.6] 24.5 [0.5–499] 16.5 [0.1–950] Data are reported as median [range] or prevalence (%) (95% confidence interval). Normal values: fT3, 2.0–4.1 pg/ml; fT4, 8.0–19.0 pg/ml; TSH, 0.35–4.50 U/l; anti-TPO = anti-thyroid peroxidase antibodies, < 250 U/l; anti-TG = anti-thyroglobulin antibodies, < 100 U/l; HTG = human thyroglobulin, < 60 ng/ml. 47 alterations in sleep and in bowel habits, pruritus, muscle cramps, paresthesias, fatigue, as well as associated somatic diseases. In this analysis, patients were classified according to their clinical status and hormone levels as having euthyroidism (n ¼ 288), hypothyroidism (n ¼ 49) and hyperthyroidism (n ¼ 31), independently of the type of thyroid disease. Results Overall HRQL was poorer in subjects with hypoand hyperthyroidism, compared with patients with goiter and thyroiditis (Figure 1). Among the domains of SF-36, role limitation-physical, role limitation-emotional, general health and social functioning were altered in all groups of patients, whereas bodily pain was never affected. In no instance did the effect size reach the conventional threshold of )0.8 indicative of a large deficit, but was in the moderate range in 25% of the cases. The domains of NHP were less severely compromised, the effect sizes being in the moderate range in only one case (social isolation in hyperthyroid patients), and being small (between 0.2 and 0.5) in approximately 15% of the cases. Unexpectedly, pain was a domain uniformly altered in all thyroid diseases when tested by means of the NHP questionnaire, whereas it was not altered in the SF-36 analysis. In both hyper and hypothyroid patients, no significant differences were observed in relation to Table 2. Effect sizes of the various domains of HRQL in the whole population of patients with thyroid diseases (mean and 95% confidence interval) Hypothyroid Hyperthyroid Goiter Thyroiditis SF-36 Physical functioning )0.46* )0.31 )0.23* )0.19 ()0.76 to )0.16) ()0.76 to 0.13) ()0.38 to )0.08) ()0.50 to 0.11) Role limitation-physical )0.63* )0.72* )0.23* )0.47* ()0.92 to )0.34) ()1.08 to )0.37) ()0.39 to )0.08) ()0.83 to )0.11) Role limitation-emotional )0.71* )0.65* )0.29* )0.47* ()0.95 to )0.46) ()1.00 to )0.30) ()0.45 to )0.13) ()0.77 to )0.16) Bodily pain )0.23 )0.02 )0.01 0.02 ()0.47 to 0.01) ()0.28 to 0.23) ()0.16 to 0.13) ()0.24 to 0.29) Vitality )0.48* )0.29 )0.36* )0.29* ()0.72 to )0.24) ()0.68 to 0.09) ()0.51 to )0.20) ()0.54 to )0.04) Mental health )029* )0.14 )0.20* )0.17 ()0.50 to )0.08) ()0.47 to 0.19) ()0.36 to )0.05) ()0.43 to 0.08) General health )0.60* )0.50* )0.38* )0.50* ()0.83 to )0.36) ()0.82 to )0.19) ()0.54 to )0.22) ()0.79 to )0.21) Social functioning )0.60* )0.41* )0.30* )0.40* ()0.90 to )0.31) ()0.76 to )0.07) ()0.46 to )0.13) ()0.66 to )0.14) NHP Energy 0.11 0.19 0.03 0.05 ()0.10 to 0.33) ()0.13 to 0.51) ()0.12 to 0.18) ()0.20 to 0.29) Emotional reactions )0.05 )0.08 )0.11 )0.14 ()0.25 to 0.15) ()0.36 to 0.20) ()0.25 to 0.03) ()0.33 to 0.14) Sleep 0.21 0.06 0.07 )0.04 ()0.041 to 0.46) ()0.27 to 0.40) ()0.08 to 0.21) ()0.27 to 0.19) Social isolation 0.21 0.63* 0.20* 0.41* ()0.02 to 0.44) (0.20 to 1.05) (0.05 to 0.35) (0.06 to 0.76) Mobility 0.32* 0.19 0.07 )0.01 (0.06 to 0.59) ()0.21 to 0.59) ()0.12 to 0.25) ()0.24 to 0.22) Pain 0.38* 0.39* 0.30* 0.29* (0.18 to 0.58) (0.16 to 0.73) (0.16 to 0.45) (0.10 to 0.49) * Significantly different from population norm, p < 0:05. 48 the severity of thyroid disease (subclinical vs. overt disease) (Figure 2), although the effect sizes indicate a poorer HRQL by a factor as large as 0.7, with the notable exception of the scales of general health in SF-36 and emotional reactions in NHP. Although in subjects with subclinical hypothyroidism all domains were no longer different from Italian population norms, most scales were significantly altered in subjects with overt disease. In hyperthyroid patients, all differences were less striking. NHP (Part II) showed that a few aspects of daily life were significantly affected by perceived health problems only in patients with hypothyroidism in comparison to the standard Italian population, after correction for age and gender (Table 3). Hypothyroid patients reported a worsening in perceived health of the previous year in 41% of the cases, no changes in 46%, and an improvement in 13% of the cases. In hyperthyroid patients, the prevalence of perceived health changes was 51, 42, and 7%. Finally, in the two euthyroid groups, the prevalence was 35, 56, and 9% for goiter and 24, 59, and 17% for thyroiditis. Most patients with hypothyroidism reported altered bowel function (41%). In particular, 33% complained of constipation, 4% diarrhoea and 4% alternate bowel pattern. The prevalence of altered bowel habits was lower in hyperthyroidism (33%, 22% constipation and 9% diarrhoea). A similar prevalence was found in those having goiter (30% had constipation, 3% diarrhoea and 4% had alternate bowel) and in thyroiditis (33, 7, 4%, respectively). Also, sleep disturbances were equally distributed among groups, being slightly more prevalent in subjects with hypothyroidism (p ¼ 0.111). Finally, mood/behaviour disturbances were equally present in all groups (from 47% in hyperthyroid to 65% in hypothyroid patients). Also, the type of alteration (i.e., anxiety, depression and panic attacks) were similarly represented, the largest prevalence being present for anxiety (over 50% of the cases), the lowest for panic attacks (10% of cases). Fatigue was reported between 7% (hyperthyroidism) and 26% (goiter) of the cases. All domains of both SF-36 and NHP questionnaires were more largely impaired in subjects with mood/behaviour disturbances, independent of the type of thyroid disease. Differences in the mean effect size between subjects with normal and Figure 1. Overall HQRL in thyroid disorders PCS and MCS of SF-36 and NHP total score. Data are presented as means and 95% confidence intervals. * p < 0:05 compared with euthyroid goiter. p < 0:05 compared with euthyroid thyroiditis. Figure 2. Domains of SF-36 and NHP questionnaires in patients with hypo- (left) and hyper-thyroidism (right), grouped based on the presence of overt (closed circles) or subclinical disease (open circles). Data are presented as the mean (95% confidence interval) of effect size, corrected for age and sex. All data crossing the zero line (population norm) are not significant. Note that in NHP, higher values represent distress. 49 altered behavior disturbances ranged from 0.28 (bodily pain, p ¼ 0:008) to 0.90 (mental health, p < 0:0001) for SF-36 and from 0.41 (social isolation, p < 0:0001) to 0.70 (physical mobility, p < 0:0001) (Figure 3). Problems associated with poor HRQL in relation to clinical status are reported in Table 4. Mood/ behavior disturbances and fatigue were significantly associated with alterations in most domains in clinically overt hypothyroidism, whereas muscle cramps and pruritus were the symptoms more commonly associated with poor HRQL in overt hyperthyroidism. In subjects with euthyroidism, the largest group being subjects with euthyroid goiter, mood/behaviour disorders (11/14 domains), fatigue (8/14 domains) and sleep problems (5/14 domains) were the reported symptoms more significantly associated with poor perceived health status. Fatigue was more commonly associated with a poor score in domains measuring physical health, whereas behaviour problems were related to altered domains measuring mental health. Discussion Our study shows that benign thyroid disease affects the perceived health status, independently of thyroid function. As hypothesized, hyperthyroid and hypothyroid patients reported poor HRQL, in Table 3. Prevalence of patients reporting problems in daily life in relation to perceived health status in thyroid disease (NHP, Part II) Problems with Hypothyroid Hyperthyroid Goiter Thyroiditis Paid employment 24 (15–35) 18 (8–32) 22 (16–29) 18 (8–31) [14 (7–24)][16 (6–29)][16 (11–22)][27 (16–42)] Jobs around the house 48 (37–60)b 42 (28–58) 35 (28–42) 27 (16–42) [29 (19–40)][29 (16–44)][33 (26–40)][47 (33–62)] Social life 16 (8–26) 33 (20–49) 21 (15–28) 14 (6–26) [16 (8–26)][20 (10–35)][20 (14–27)][29 (17–44)] Home life 34 (24–46)a 27 (15–42) 26 (19–33) 14 (6–26) [18 (10–28)][18 (8–32)][20 (14–27)][31 (19–46)] Sex life 29 (19–40)a 22 (11–37) 20 (14–27) 12 (4–24) [14 (7–24)][16 (6–29)][17 (12–23)][27 (16–42)] Interests and Hobbies 30 (20–42)b 29 (16–44) 20 (14–26) 18 (8–31) [13 (6–23)][16 (6–29)][16 (11–22)][25 (14–40)] Holidays 25 (16–36) 20 (10–35) 23 (17–30) 18 (8–31) [17 (9–27)][18 (8–32)][20 (14–26)][27 (16–42)] Data are presented as % (95% confidence intervals). The prevalence in normative Italian population, corrected for gender and age, is reported in squared brackets. Values in patients with hypothyroidism are different from normative population, after correction for age and gender. a p < 0:05; b p < 0:025. Figure 3. Domains of SF-36 and NHP questionnaires in patients with thyroid disorders, grouped based on the presence (closed circles) or absence (open circles) of self-reported mood/ behaviour disorders. Data are presented as the mean (95% confidence interval) of effect size, corrected for age and sex. 50 agreement with previous studies [5, 6, 8, 10]. An unexpected finding was that HRQL was poor also in subjects with euthyroid thyroiditis and in the large group of patients with euthyroid goiter, when compared with the general Italian population after adjustment for gender and age. Apparently, not only clinically overt thyroid diseases, but simply being labelled as having a thyroid disturbance, alters the perceived health status. We used effect sizes to compare the extent of HRQL impairment to standard Italian norms. For both questionnaires, control values were obtained in random samples of the Italian population. This procedure has been previously used in the study of HRQL in different populations (liver disease [18, 19], obesity ), and to identify the role of symptoms in poor perceived health status. The benchmarks used to define the extent of the effect size have been supported by clinical studies . This kind of representation has two additional advantages. First, it standardizes data for sex and age, and further corrections for these two relevant factors are no longer needed. Secondly, it gives an immediate picture as to how much and how significantly a specific disease affects HRQL in comparison to the standard HRQL of the general population. The two questionnaires we used have been extensively applied in epidemiological studies to compare the impact of chronic diseases in the general population and to determine health policies and resource allocation [21–23]. Although they partly overlap, their conceptual models differ. SF36 identifies both positive and negative aspects, and is intended to measure the full range of health status and well-being, both physical and mental, Table 4. Factors significantly associated with poor HRQL in relation to clinical status in subjects with thyroid disorders (odds ratio and 95% confidence interval) Euthyroid N = 288 Hypothyroid N = 49 Hyperthyroid N = 31 SF-36 Physical functioning Paresthesias (2.8, 1.6–5.0) Sleep (2.0, 1.1–3.6) Fatigue (9.0, 2.0–40.8) – Role limitation-physical Fatigue (3.0, 1.7–5.3) Bowel habits (1.9, 1.1–3.2) Fatigue (8.3, 1.8–38.1) Pruritus (6.5, 1.3–33.0) Role limitation-emotional Behavior (2.3, 1.4–3.9) Fatigue (2.2, 1.2–3.9) Behavior (3.5, 1.0–12.1) Pruritus (6.9, 1.2–40.4) Bodily pain Behavior (1.9, 1.1–3.3) – Cramps (17.2, 1.5–210.2) Vitality Fatigue (3.0, 1.7–5.5) Behavior (2.5, 1.4–4.4) Fatigue (16.3, 2.3–112.7) Behavior (7.7, 1.4–42.9) Behavior (6.5, 1.3–33.0) Mental health Behavior (5.6, 2.7–11.6) Fatigue (1.9, 1.0–3.6) Behavior (14.2, 1.7–121.8) Cramps (9.5, 1.4–64.4) General health Behavior (4.6, 2.5–8.5) Paresthesias (1.8, 1.1–3.1) Behavior (11.6, 1.4–100.2) Sleep (10.5, 1.7–63.9) Social functioning Behavior (3.2, 1.8–5.7) Paresthesias (1.8, 1.0–3.4) Behavior (30.4, 3.5–263.8) – NHP Energy Sleep (10.0, 4.4–23.0) – Cramps (7.5, 1.1–49.3) Emotional reactions Sleep (3.2, 1.5–7.1) Behavior (3.0, 1.3–6.8) Paresthesias (12.9, 1.4–118.6) Cramps (8.2, 1.0–66.2) Sleep Behavior (4.9, 2.1–11.5) Fatigue (3.0, 1.5–6.0) Sleep (2.5, 1.2–5.1) Pruritus (4.8, 1.1–22.3) Sleep (10.0, 1.0–97.1) Social isolation Behavior (2.8, 1.5–5.1) Fatigue (2.2, 1.2–4.0) Behavior (9.0, 1.0–78.2) Behavior (8.0, 1.3–48.2) Mobility Behavior (2.5, 1.3–4.8) Fatigue (2.4, 1.3–4.4) – Sleep (8.0, 1.3–48.2) Pain Behavior (3.7, 1.9–7.2) Sleep (2.2, 1.2–4.0) Fatigue (2.1, 1.1–3.9) – Behavior (6.1, 1.0–36.9) The list of independent variables considered in the analysis is reported in the Materials and methods section. 51 also including the personal evaluation of health. Accordingly, the results may be used for comparison between different populations, between different specific treatments and across different health care delivery systems. By contrast, NHP only measures negative aspects [24, 25]. Therefore, the questionnaire is very specific in detecting any change in the individual’s perception of physical and emotional problems affecting his/her daily living, but it is less sensitive in normal persons to measure how much healthy they feel, because of a large ‘floor effect’. Accordingly, although the majority of domains of SF-36 were impaired, NHP domains were less affected. The combined use of the two questionnaires and the agreement of results give further credit to the analysis presented here. In addition, the two instruments allow a more comprehensive evaluation of HRQL, covering aspects not shared by a single questionnaire. Apparently, thyroid diseases have a large impact on domains related to both the physical and the mental dimensions of health status. Consequently, the larger impact is observed on a domain that shares both dimensions, namely general health. This last scale is dependent on both severe and minor symptoms, irrespective of their association with clinical disability . A relatively pure scale of physical dimension, such as physical functioning , or a pure scale of mental dimension as mental health, was less affected. This also applies to NHP, where the Energy domain, a pure physical dimension, was not significantly compromised. The two scales measuring the physical dimension of pain deserve a specific comment. Surprisingly, bodily pain of SF-36 was not significantly impaired, whereas the scores of pain scale of NHP were poor. This implies that the two scales measure different aspects of pain, and that the NHP scale more precisely picks up the distress generated by disease. Bodily pain of SF-36 was previously shown to be associated with the physical dimension of health more than with the mental dimension ; this might not be the case for pain of NHP. In spite of a reduced HRQL, only hypothyroid patients reported an increased prevalence of problems associated with everyday life, namely jobs around the house, home life and interests and hobbies. Although hyperthyroid patients reported a similarly increased prevalence of perceived problems, the smaller number of patients may be a reason for a type II error. This specific question needs to be further addressed. The most relevant result of the present study is that a poor HRQL was not limited to subjects with overtly symptomatic thyroid disease, either hypoor hyperthyroidism, but was also present in asymptomatic diseases, such as euthyroid goiter or Hashimoto thyroiditis. HQRL in these last groups was also poorer than in patients with subclinical hypo- and hyperthyroidism, whose quality of life was in the range of the control population. Previous studies have convincingly related poor HRQL with clinically significant symptoms in hyperthyroidism [7, 8]. This was confirmed in the present analysis, where fatigue and behavior disorders (mainly depression), severely disturbing hypothyroid patients, were associated with poor scores in several domains. This is also the case of hyperthyroid patients, where muscle cramps and pruritus may contribute to the burden of disease, as much as they affect the quality of life in different clinical conditions . The poor HRQL in euthyroid thyroid disease is more difficult to explain. It gives rise to the question as to whether these patients are really asymptomatic. Poor HRQL might be solely due to awareness of a disease state, as demonstrated in other disease states . Alternatively, undetectable symptoms which are not diagnosed by the physicians but are perceived better by patients may be present, generating the distress measured by HRQL. In both cases an underlying alteration in mood/behaviour, recognized as a definite psychopathological disease only in a minority of patients, identifies subjects at risk of having a poor HRQL. In agreement with this conclusion, a reduced well-being was reported also in subjects with subclinical or remitted hyperthyroidism [10, 29] and partly confirmed in the present study, psychopathological features were also present in subclinical hypothyroidism , and higher rates of frank psychiatric disorders were present in comparison to the normal population . The finding that thyroid patients without mood/behaviour disturbances report a better HRQL than the population norm (Figure 2) underlines the importance of psychopathology in perceived health status in all conditions. Had normative values also been derived from a selected population without mood/behavior disorders, HRQL scores would be much better. 52 In conclusion, patients with thyroid diseases, independent of the severity of thyroid dysfunction, perceived diminished health status. In their clinical evaluation, physicians need to consider this aspect for a more comprehensive care of patients , as recently supported by regulatory agencies .