Journal reference: Rondondi N, den Elzen WPJ, Bauer DC, et al. Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA. 2010;304(12):1365-1374 ^[1]

Link: http://dx.doi.org/10.1001/jama.2010.1361

Evidence cookie says...

Subclinical hypothyroidism (elevated TSH with normal thyroxine levels) is associated with increased risk of coronary heart disease (CHD) events and mortality. TSH levels ≥ 7.0 mIU/L → trend for increasing risk. TSH level up to 6.9 mIU/L → no increased risk It is reasonable that asymptomatic subclinical hypothyroidism with minimal increases of TSH up to 6.9 mIU/L not require treatment with thyroxine. However, testing for thyroid antibodies and TSH monitoring is recommended to detect conversion to overt hypothyroidism.

More details:

Article details

Study design:

(participant-level) meta-analysis of prospective cohort studies

Study aim:

assess the risk of CHD and total mortality for adults with subclinial hypothyroidism

Methods summary:

• systematic literature search, criteria developed a priori; inclusion criteria (full-text, longitudinal cohort studies, measured TSH and T4 level at baseline, followed patients systematically over time, assessed CHD events and/or mortality, had a euthyroid comparison group)
• invitation to investigators of eligible stiudies to join collaboration to collect participant level detail (including demographic details for traditional cardiovascular risks)
• outcome measures were CHD events, CHD mortality and total mortality
• primary analyses were adjusted for age and sex and cardiovascular risk factors

Results summary:

• 12 prospective studies met eligibility criteria and 11 studies agreed to provide individual participant data
• 55,287 adults
• Although there was a trend to increasing risk of CHD events, CHD mortality and total mortality when comparing those participants with subclinical hypothyroidism to euthyroid, it was not statistically significant (95% confidence intervals for all three include 1.0)
• However, analysis of subgroups strongly demonstrated increasing risk of CHD events (P < 0.001 for trend) and CHD deaths (P = 0.005 for trend).

In age and sex adjusted analyses, hazard ratio for CHD events for the following TSH levels (mIU/L):

• 4.5 to 6.9: 1.00 (95% CI, 0.86-1.18)
• 7.0 to 9.9: 1.17 (95% CI, 0.96-1.43) (note: CI includes 1.0)
• 10.0 to 19.9: 1.89 (95% CI, 1.28-2.80)

Hazard ratio for CHD mortality for the following TSH levels (mIU/L):

• 4.5 to 6.9: 1.09 (95% CI, 0.91-1.30)
• 7.0 to 9.9: 1.42 (95% CI, 1.03-1.95)
• 10.0 to 19.9: 1.58 (95% CI, 1.10-2.27)

Study conclusion:

Subclinical hypothyroidism is associated with an increased risk of CHD events and CHD mortality in those with higher TSH levels, particularly those with a TSH concentration of 10 mIU/L or greater.

Participants:

• 55,287 adults with 3,450 with subclinical hypothyroidism
• mix of ages with the two largest studies (HUNT Study and EPIC-Norfolk Study) that have over half the participants having median agest of 55 years (41-98 interquartile range) and 58 years (39-78 IQR) respectively
• majority of participants at baseline and follow up not on thyroid medications
• predominately white population

Methodological weaknesses

• population predominately white
• reported results are adjusted with sex and age only; it is claimed that further adjustment for traditionally cardiovascular risk factors yielded similar results but the results are not explicitly given
• relatively wide confidence intervals for results demonstrating some limitations of statistical power
  • specifically, the study was powered to detect a HR of 1.18 or higher for CHD events and 1.30 or higher for CHD mortality and 1:13 or higher for total mortality
  • it is very likely that the study was underpowered comparing euthyroid and subclinical hypothyroid groups as a whole
• moderate level heterogeneity remained

Methodological strengths

• participant level meta-analysis
• rigorous attempt at reducing heterogeneity and this was measured
• large study population with ages relevant to subclinical hypothyroidism seen in primary health care
• many studies included a long follow up period (median years follow up from 2.5 to 20.0)
• power calculation is given
• primary analysis have been adjusted for age and sex and (claimed) traditional cardiovascular risk factors (SBP, smoking, total cholesterol, diabetes)
• the conclusion appears to be valid

Biases and conflicts of interests

• Nil declared and nil obvious.
• Note: included studies in the meta-analysis were not individually reviewed

Clinical relevance to primary health care

Subclinical hypothyroidism is common. By age sixty, the population level prevalence is approximately 10% ^{[2], [3]}. Coronary heart disease risk in subclinical hypothyroidism has been a concern. Management of subclinical hypothyroidism with TSH levels below 10 mIU/L has been controversial with some authors arguing for routine treatment with thyroxine and others recommending selective treatment ^[2].

This study provides some guidance; patients with subclinical hypothyroidism with a TSH of 6.9 mIU/L or less do not appear to be at increased risk of cardiovascular outcomes. Patients with a TSH of 7.0 mIU/L or greater should be considered for treatment.

It should be noted that this study does not provide any evidence on whether treatment of subclinical hypothyroidism with thyroxine reduces cardiovascular risk. A meta-analysis of 350 patients in randomised control trials did not find improvement in survival or cardiovascular morbidity^[4]. However, the external validity of this result is hampered by the small number of patients and short follow up duration of the trials (6 to 14 months).

References

1. Rondondi N, den Elzen WPJ, Bauer DC, et al. Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA. 2010;304(12):1365-1374.
2. Fatourechi V. Subclinical hypothyroidism: an update for primary care physicians. Mayo Clinic Proceedings January 2009, vol.84 no.1 p65-71
3. Managing subclinical hypothyroidism [reprint from Drug and Therapeutics Bulletin 1998;36:1-3]. Aust Prescr 1999;22:132-4
4. Villar HCCE, Saconato H, Valente O, Atallah ÁN. Thyroid hormone replacement for subclinical hypothyroidism. Cochrane Database of Systematic Reviews 2007, Issue 3. Art. No.: CD003419.