Feb 25

Incidence and progression of thyroid dysfunction in elderly

Journal reference: Gopinath B, Wang JJ, Kifley A, et al. Five-year incidence and progression of thyroid dysfunction in an older population. Internal Medicine Journal, 2010;40(9):642–649 [1]

Link: http://dx.doi.org/10.1111/j.1445-5994.2009.02156.x

Published: 15 September 2010

Evidence cookie says...

The rate of progression from subclinical hypothyroidism to overt hypothyroidism is about 1 in 6 for people over five years for those aged 55 years or greater.

Female gender, obesity and a TSH > 2.0 mIU/L were all associated with an increased 5-year incidence of thyroid dysfunction.

Note: caution should be taken to avoid over-generalising this data; in itself it does not support any specific management strategy.

More details:

Article details

Study design:

population-based prospective cohort study

Study aim:

assess 5-year incidence, progression and risk factors for development of thyroid dysfunction in a single older population

Methods summary:

  • Data from The Blue Mountains Eye Study (original population cohort study) was analysed
    • Original study (looking at age-related eye diseases) recruited Australians aged 49 years and older from two postcode areas in the Blue Mountains
    • 3654 of 4433 (82.4%) eligible residents were examined
  • thyroid function status was assessed at follow up examinations at year 5 and 10
  • 1768 participants (75.7% of the 2334 survivors at 5-years from the original baseline study) ≥ 55 years had thyroid function assessed
  • Patients receiving any form of treatment for their thyroid condition at baseline were excluded

Results summary:

  • 5-year incidence of thyroid dysfunction was 4.7% (CI 3.4-6.1)
  • 5-year incidence of subclinical hypothyroidism was significantly higher in women than in men, 2.5% vs 0.7% (P = 0.03)
  • The highest incidence of thyroid dysfunction was in those aged > 80 (8.8%)
  • Progression from subclinical to overt hypothyroidism was observed in 17.9% of subjects
  • Predictors of hypothyroidism over 5-years (OR = odds ratio):
    • female gender (OR 3.01, CI 1.42-6.40)
    • serum TSH > 2 mIU/L:
      • overt hypothyroidism (OR 4.11, CI 1.61-10.51)
      • subclinical hypothyroidism (OR 11.74, CI 1.48-92.90)
    • obesity after adjustment for sex and TSH (OR 2.35, CI 1.14-4.81)

Study conclusion:

The 5-year incidence of thyroid dysfunction in this older population was relatively low, and was associated with obesity and serum TSH level > 2 mIU/L at baseline. Over one in six persons with subclinical hypothyroidism progressed to overt thyroid dysfunction over the 5-year period. Our findings highlight the need for appropriate management of subclinical hypothyroidism among older people.


  • 82.4% of eligible residents in two post codes of the Blue Mountains
  • however, just under half of the original cohort (48%) had repeat thyroid function tests performed at 5-years (of the survivors of the original cohort, 24.3% did not have thyroid function tested at this time)
  • non-participants were more likely to be older than participants (P < 0.0001) and more likely to be male (P = 0.01)

Methodological weaknesses

  • There are some inherent weaknesses in an evaluation of data from a study with a different primary goal.
    • In this case, the cohort of patients who had thyroid function tests performed at year 5 and 10 had substantially diminished compared to the original cohort.
    • There is the possibility of significant biases resulting from differences between the partipants and those who were otherwise eligible but dropped out.
  • Some participant demographic detail is unreported though appears to have been measured in the original study, e.g., medical conditions, medications, lifestyle factors, etc.
    • Without this information, the validity of generalisation of the findings to other populations is problematic.
  • Subgroups of participants (particularly those who developed thyroid dysfunction) had small numbers.
  • This is reflected by the very wide confidence intervals for the odds ratios for predictors of thyroid dysfunction.

Methodological strengths

  • Original study recruited a large proportion of all eligible candidates in a geographic area.

Biases and conflicts of interests

Nil declared and none seem obvious.

Clinical relevance to primary health care

The fact that the conversion rate from subclinical hypothyroidism to overt hypothyroidism of about 18% over five years in this older population is interesting. Nevertheless, caution is required due to the (very) small number of participants to whom this occurred in the study and the absence of a confidence interval.

Many of the findings in this study support suspected or known risks for developing hypothyroidism; female gender, obesity and an elevated TSH. Potential biases in the selection of the participants, the low numbers of participants in subgroups and the wide confidence intervals in some of the results should dissuade us from making any confident assumptions about the numerical accuracy of the detected predictors of hypothyroidism. In itself, this study does not support any specific strategy for screening or management of any form of thyroid dysfunction in the elderly.

The discussion on possible reasons why the incidence of hypothyroidism in this cohort seemed low should be considered speculation and hypothesis generating only rather than factual.


  1. Gopinath B, Wang JJ, Kifley A, et al. Five-year incidence and progression of thyroid dysfunction in an older population. Internal Medicine Journal, 2010;40(9):642–649
Print Friendly, PDF & Email


    • Joel Rhee on 25 February 2011 at 11:05 AM
    • Reply

    I haven’t yet read the article but I wonder whether they checked a person’s thyroid autoantibody status to see whether this was predictive of progression to overt hypothyroidism…

    1. This was not done in this study but I’m pretty sure that there is pre-existing evidence that demonstrates that autoantibodies are predictive.

Leave a Reply

%d bloggers like this: