1. Subclinical thyroid disorders
Mario Skugor M.D. FACE
Associate Professor of Medicine
CCLCM of CWRU
Cleveland Clinic

2. Definitions:
Individuals with elevation of TSH but normal thyroid hormone levels have sub-clinical hypothyroidism*.
*Some authors consider patients with high-normal TSH and abnormal response to TRH as having SCH.
Gharib H. et al. J Clin Endocrinol Metab 2005;90:

3. Definitions:
Individuals with TSH lower than normal but with normal thyroid hormone levels have sub-clinical hyperthyroidism.
Gharib H. et al. J Clin Endocrinol Metab 2005;90:

4. Normal TSH??? Normal TSH range Most common assays in U.S.A.
uU/mL
uU/mL

5. Normal TSH???
Lower limit of the normal range is somewhat variable but relatively well established.

6. Normal TSH???
Upper level of normal range is controversial.

7. Normal TSH???
0.4
5.5

8. NHANES III N = 13,344 Healthy, antibody negative subjects median 1.39
0.45
4.12

9. Normal TSH???
Consensus conference (ATA, AACE, Endo Soc) accepted NHANES III derived normal range.
National Academy of Clinical Biochemistry proposed uU/mL for normal range.
AACE proposed uU/mL

10. Practical considerations
Results of lowering upper limit of normal range from 5 to 2.5 uU/mL changes prevalence of sub-clinical hypothyroidism:
4.6% to 20%
Fatourechi V. et al. JAMA 2003;290:

11. Practical considerations
TSH varies with time.
Mildly abnormal TSH should be rechecked in 3-4 months to confirm persistent elevation.
Transient elevations are not uncommon.

12. Practical considerations
TSH elevations NOT associated with SC hypothyroidism:
Recovery from non-thyroidal illness.
Adrenal insufficiency.
Treatment with metoclopramide and domperidone
TSH secreting adenomas
Thyroid hormone resistance
Assay variability
Late evening pulse in TSH secretion

13. Practical considerations
Decreased TSH also could be seen with conditions that are not associated with hyperthyroidism.

14. Practical considerations
Decreased TSH NOT associated with SC hyperthyroidism:
Central hypothyroidism
Non-thyroidal Illness
Glucocorticoid and dopamine use
Recovery from hyperthyroidism
Early pregnancy

15. Practical considerations
Even TSH over 20 uU/mL can be associated with normal TH levels.

16. Practical considerations
Attempts to identify individuals with mild thyroid disorders based on search for thyroid specific symptoms have not been successful.

17. Sub-clinical hypothyroidism
Should be diagnosed in patients who do NOT complain of symptoms associated with hypothyroidism.
In patients who are screened for presence of thyroid disorder because of increased risk.

18. Elevated TSH normal TH
If patient complains of symptoms and elevated TSH with normal TH is found, than disorder cannot be considered to be sub-clinical.

19. Mild hypothyroidism
In my opinion, after rechecking to confirm TSH elevation, treatment should be seriously considered in such patients.
TSH should be corrected to below 2.0 uU/mL and therapeutic effects assessed after 2-3 months.
If benefit is derived treatment should be continued.

20. Sub-clinical hypothyroidism
Individuals with TSH from upper limit of normal to 10 uU/mL
With normal TH levels
With no symptoms that could be explained by hypothyroidism
Have “Sub-clinical hypothyroidism”

21. Epidemiology Whickham survey Prevalence Men 2.8% Women 7.5%
Women >60 y 11.6%
Vanderpump MP. Et al. Clin endocrinol (Oxf) 1991;43:55-69

22. Epidemiology NHANES III Prevalence in 16,533 individuals
Patients with known thyroid disease were excluded.
SCH found in 4.3%
Hollowell JG, et al. J Clin Endocrinol Metab 2002;87:489.

23. Epidemiology Colorado Thyroid Disease Prevalence Study N – 25,862
Overall 9.5%
18-24 y 4.0%
Women >74 22%
Men >74 y 16%
Canaris GJ. Et al. Arch Intern Med 2000:160: /

24. Epidemiology Increased prevalence is reported in patients with:
Diabetes mellitus type 1 (1)
Schmidt’s syndrome (2)
Celiac disease (3)
Due to presence of Hashimoto’s thyroiditis
1. Vondra K. et al. J Endocrinol Invest 2004: 27:
2. Betterie C. et al. Endocr rev 2002;23:
3. Sategna-Guidetti C. et al. Am J Gastroenterol 2001:96:

25. Etiology–SC hypothyroidism
Common causes
Iatrogenic
In patients with coexisting heart disease
In patients with malignant diseases
In patients with poor compliance
In patients taking medications inappropriately

26. Etiology–SC hypothyroidism
Common causes
Iatrogenic
Hashimoto’s thyroiditis
Most common cause
54% of SCH patients in U.S. are abs positive (1)
UK study found abs in 67% of women and 40% of men with SCH (2)
Hamburger JI, et al. NEJM 1985;313:267.
Tonbridge WM. Et al. Br Med J 1974;3:89.

27. Etiology–SC hypothyroidism
Common causes
Iatrogenic
Hashimoto’s thyroiditis
Postpartum thyroiditis

28. Etiology–SC hypothyroidism
Common causes
Iatrogenic
Hashimoto’s thyroiditis
Postpartum thyroiditis
Medications
Amiodarone
Lithium
Interferon
Sunitinib
Sorafenib

29. Etiology–SC hypothyroidism
Common causes
Iatrogenic
Hashimoto’s thyroiditis
Postpartum thyroiditis
Medications
Partial thyroidectomy

30. Etiology–SC hypothyroidism
Common causes
Iatrogenic
Hashimoto’s thyroiditis
Postpartum thyroiditis
Medications
Partial thyroidectomy
Head and neck radiation

31. Etiology–SC hypothyroidism
Common causes
Iatrogenic
Hashimoto’s thyroiditis
Postpartum thyroiditis
Medications
Partial thyroidectomy
Head and neck radiation
RAI ablation of Graves’ disease

32. Clinical consequences
Progression to overt hypothyroidism
2-5% yearly
10% may eventually normalize
Higher initial TSH – higher the risk of progression
Positive AMA-s – higher risk of progression
Huber G. et al. J Clin Endocrinol Metab 2002;87:
Diez JJ. et al. J Clin Endocrinol Metab 1991:72:

33. Clinical consequences
Development of hypothyroidism, 20 year follow up – Whickham survey
27% if abs positive
33% if TSH higher than normal
55% if TSH higher and abs positive
4% if TSH was normal and abs negative
Vanderpump MP. Et al. Clin endocrinol (Oxf) 1991;43:55-69

34. Clinical consequences
Vanderpump MP. Et al. Clin endocrinol (Oxf) 1991;43:55-69

35. Clinical consequences
Risk of CV disease
Diastolic dysfunction
Diastolic hypertension
Increase in LDL-C
Increased hsCRP
Alteration in coagulation parameters
Endothelial dysfunction
Alber CP. et al. Am Heart J 1964
Diez JJ. et al. J Clin Endocrinol Metab 1991:72:

36. Clinical consequences
Risk of CV disease
Whickham survey showed no association over 20 years with:
Coronary heart disease
Dyslipidemia
Mortality
Vanderpump MP, et al. Clin Endocrinol (Oxf) 1991:43:55-69.

37. Clinical consequences
Risk of CV disease
Rotterdam study showed increased prevalence of atherosclerosis (odds ratio 1.7) and MI (odds ratio 2.3) in women with SCH.
Hak AE, et al. Ann Intern Med 2000;132:

38. Clinical consequences
Risk of CV disease
Studies show mixed results but population studied are different and number of patients are usually small (total number of subjects is high but patients with SCH are not many).
Example – 4331 subjects but 281 with SCH (1)
Pearce EN, et al. Endo Soc, Boston MA abstract 42-1.

39. Controversy
McDermott MT, Ridgway EC 2001 Subclinical hypothyroidism is mild thyroid failure and should be treated. J Clin Endocrinol Metab 86:4585–4590.
Chu JW, Crapo LM 2001 The treatment of subclinical hypothyroidism is seldom necessary. J Clin Endocrinol Metab 86: 4591–4599.

40. Clinical Consequences
Fertility
17% of Tg and AM antibodies positive had spontaneous miscarriage, compared to 8.4% of antibody negative women (1)
2.3% of 25,756 pregnant women had SCH, and their risk for placental abruption and preterm delivery was 2 times normal (2)
1. Stagnaro-Green A, et al. JAMA 1990;264:
2. Casey BM, et al.Obstet Gynecol 2005;105:

41. Clinical Consequences
Fertility
The 16 and 11 recurrent aborters with positive TPO-abs were treated with T4 and IV immunoglobulins respectively.
T4 treated patients had 81.2% delivery rate and IV Immunoglobulin treated 54.5%.
Vaquero E, et al. Am J Reprod Immunol 2000;43:

42. Clinical Consequences
Neuropsychiatric disorders
SCH patients with depression have poorer response to antidepressant medications (1)
SCH patient have increased lifetime prevalence of depression (2)
Women with SCH and goiter have increased rate of anxiety and depressive features (3)
Joffe RT, Levitt AJ. Psychoneuroendocrinology 1992;17:215
Haggerty JJ, et al. Am J Psychiatry 1993:150:508.
Monzani F, et al. Clin Invest 1993;71:367.

43. Clinical Consequences
Mortality
CV health study (N = 3,233) showed NO increase (1)
Meta-analysis showed increased risk in those older than 65 (2)
Several studies showed increase in CV and all-cause mortality (3-5)
Cappola AR, et al. JAMA 2006;295:1033.
Razvi S, et al. J Clin Endocrinol Metab 2008;93:2998.
Imaizumi M, at al. J Clin Endocrinol Metab 2004:89:3365.
Iervasi G, et al. Arch Intern Med 2007;167:1526.
Walsh JP, et al. Arch Intern Med 2005;165:2467.

44. Our experience
We examined cohort of 6,408 patients at HIGH RISK for cardiovascular disease.
Patients were enrolled in cardiovascular disease prevention program of Cleveland Clinic between January 16, 1995 and June 25, 2008.
Data about patients were entered in the Preventive Cardiology Information System (PreCIS) database.
McQuade C, et al. ATA meeting, Chicago IL, 2008 abstract 311.

45. Methods Patients: Secondary prevention ~ 50% Primary prevention ~ 50%
Hypercholesterolemia
Strong family history of CVD
The CVD risk factors are aggressively treated.

46. Methods (cont) Five TSH intervals predefined:
<0.4 µU/ml (hyperthyroid)
µU/ml (normal)
µU/ml (mild SCH)
µU/ml ( moderate SCH)
>10.0 µU/ml (hypothyroid)

47. Methods (cont) Number of patients: <0.4 µU/ml 168 patients
No TSH available 14 patients

48. Patients characteristics:
There was no difference between groups:
Framingham risk score
Diabetes mellitus
LDL
Lp(a)
Microalbuminuria
Body mass index
Diastolic blood pressure

49. Results Adjustment was made for: Age Sex Smoking HDL LDL
Systolic blood pressure
Diastolic blood pressure
Diabetes

50. Results
After adjusting for common risk factors there were no increased risk for the group of patients with TSH less than 0.4.

51. All cause mortality - adjusted
All Patients (n = 6,394)
p<0.001
P<0.001
Source: Arial 9pt, left-aligned with slide title

52. All cause mortality - adjusted
All Patients (n = 6,394)
P=0.021
Source: Arial 9pt, left-aligned with slide title

53. Limitations
Thyroid status based only on elevated TSH values as L-T4 and T3 values are not consistently available
The PreCIS database consists of exclusively preventive cardiology patients at the Cleveland Clinic. The cohort is highly selected for patients with or at risk for CHD. Results may not be generalizable
Cause of death cannot be determined from the SS death index.

54. Effect of treatment
Psychometric outcomes and hypothyroid symptoms scores in double blind randomized trials:
Improvement is seen consistently in those with TSH >10 uU/mL.
Cooper DS, et al. Ann Intern Med 1984;101:18.
Nystrom E, et al. Clin Endocrinol 1988;29:63.
Jaeschke R, et al. J Gen Intern Med 1996;11:744.
Meier C, et al. J Clin Endocrinol Metab 2001;86:4860.

55. Effect of treatment Cholesterol T4 treatment decreases:
Total Cholesterol in those with initial level over 240 mmol/L (meta-analysis)
LDL – C
Apolipoprotein B-100
Danese MD, et al. J Clin Endocrinol Metab 2000;85:2993.
Mikhail GS, et al. Endocr Pract 2008;14:570.
Iqbal A, et al. J Intern Med 2006;260:53.

56. Effect of treatment Cardiac function, blood pressure
Cardiac output increases (1)
Vascular resistance decreases (1)
Mean arterial pressure decreases (1)
Carotid intimal-media thickness decreases (2)
Diastolic and systolic function improve (3)
1. Faber J, et al. Thyroid 2002;12:319.
2. Adress M, et al. Clin Endocrinol 2008.
3. Turhan S, et al. J CLin Endocrinol Metab. 2006;91:3490.

57. Bottom line:
I treat individuals with SCH if there are associated clinical symptoms or positive thyroid autoantibodies.
A 2-6 months later clinical and laboratory response is assessed.
Decision to treat or not to treat further is made in discussion with patient.

58. SC hyperthyroidism Etiology Iatrogenic
25% of patients taking T4 may have low TSH.
Some of these are intentional
Thyroid carcinoma
History of head and neck radiation
Presence of nodules and goiters
De Whalley P. Br J Gen Pract 1995;45:93.

59. SC hyperthyroidism Etiology Iatrogenic
Autonomous nodules (solitary and in MNG)
A 22% of patient with MNG found to have SC hyperthyroidism (1)
In hyperthyroid patients over 55 years of age with MNG 57% were sub-clinical (2)
Diaz JJ. Gerontology 2003;49:316
Rieu M, et al. Clin Endocrinol 1993;39:67

60. SC hyperthyroidism Etiology Iatrogenic
Autonomous nodules (solitary and in MNG)
Graves’ disease
About 6% of patients with Graves’ disease are sub-clinical.
Diaz JJ. Gerontology 2003;49:316.

61. SC hyperthyroidism Etiology Iatrogenic
Autonomous nodules (solitary and in MNG)
Graves’ disease
Thyroiditis
Graves’ disease in remission
High hCG

62. SC hyperthyroidism Epidemiology Reports vary from 0.7% to 12.4%
NHANES III showed 0.7% (1)
Smokers two times higher prevalence.
Another study found 1.8% with TSH <0.1 uU/mL in U.S. (2)
In England 6% had TSH <0.5 uU/mL (3)
Hollowell JG, et al. J Clin endocrinol Metab 2002;87:489
Sawin CT, et al. Arch Intern Med 1991;151:165.
Parle JV, et al. Clin Endocrinol (Oxf) 1991;34:77.

63. SC hyperthyroidism Natural course 40-60% normalize (1-3)
About 4% progresses to overt hyperthyroidism (4,5)
Eggersten R, et al. BMJ 1988;297:1586
Parle JV, et al. Clin Endocrinol (Oxf) 1991;34:77.
Meyerovitch J et al. Arch Intern Med 2007;167:1533.
Sawin Ct, et al. Arch Intern Med 1991;151:165
Sandrock D, et al. Acta endocrinol (Copenh) 1993;128:51

64. Clinical consequences
Bone metabolism
No good data on fracture rates.
A meta-analysis found decreased BMD, but only in postmenopausal women (1).
In those taking T4 low BMD was found only if dose was higher than 1.6 mcg/kg/d (2).
Markers of bone turnover have been reported to be increased (3)
Faber J, Galloe AM. Eur J Endocrinol 1994;130:350
Schneider DL, et al. JAMA 1994;271:1245
Loviselli A, etr al. Thyroid 1997;7:561

65. Clinical consequences
CV effects
Atrial fibrillation
Prospective study- 2,000 adults over 60 followed for 10 years
TSH <0.1 uU/mL % cumulative incidence
TSH uU/mL - 16% cumulative incidence
TSH >0.4 uU/mL - 11% cumulative incidence
Cappola AR, et al. JAMA 2006;295:1033.

66. Clinical consequences
CV effects
Atrial fibrillation
A cross-sectional study, 24,000 hospitalized patients.
14% prevalence if overtly hyperthyroid
13% prevalence in SC hyperthyroidism
2% prevalence in euthyroid controls
Auer J, et al. Am Heart J 2001;142:838

67. Clinical consequences
CV effects
Other described effects
Increased heart rate and atrial premature beats
Reduced exercise tolerance
Increased cardiac contractility
Increase in LV mass index
Lower LDL-C and total cholesterol
Systolic hypertension

68. Clinical consequences
Mortality
In 3,121 cardiac patients SC hyperthyroidism was associated with 2.32 increased risk for cardiac death (1)
Meta-analysis of 290 subjects over 60 found 41% increase in all cause mortality (2)
Cardiovascular Health Study (3,233 subjects over 65 y) found no increase in mortality (3)
Iervasi G, et al. Arch Intern Med 2007;167:1526
Haentjens P, et al. Eur J Endocrinol 2008;159:329.
Cappola AR, et al. JAMA2006;295:1033

69. SC Hyperthyroidism
Most individuals with TSH of 0.1 um/mL or higher do not experience any symptoms.

70. Management High risk patients Low risk patients
TSH <0.1 uU/mL evaluate and treat.
TSH uU/mL evaluate and treat if BMD low, a-fibb or symptoms of hyperthyroidism are present.
Low risk patients
TSH <0.1 uU/mL evaluate and treat if BMD low or symptoms present.
TSH uU/mL observe.

71. Questions What is the appropriate normal range for TSH?
Answer will require more studies that will need to include antibody determination and more than one determination of TSH over period of time to avoid transient elevations to creep into the normal ranges.

72. Questions Should we screen for SC thyroid dysfunction?
Population screening is probably not cost effective.
Testing patients with mild symptoms suggestive of hypothyroidism is NOT screening.

73. Questions Who should we screen for SC thyroid dysfunction?
High risk groups should be screened.
History of neck irradiation
Pituitary surgery or irradiation
Use of medications known to affect thyroid
Presence of thyroid nodule or MNG

74. Questions Should we screen for SC thyroid dysfunction?
Studies needed about
Patients with autoimmune disorders
Unexplained infertility
Previous postpartum thyroiditis
Refractory depression
Turner’s and Down’s syndrome
Obesity

75. Questions Is there a treatment benefit? Studies are inconclusive.
Patients with mild laboratory abnormalities should not be treated automatically.
These should be followed in 3-6 months and reassessed.
If even mild symptoms are present treatment should be offered.
Long term outcomes need studies