Serum thyroglobulin in hospitalized chronic geriatric patients: its relationship to age, non-thyroidal illness, goitre and thyroid dysfunction in a follow-up study.

I. Szabolcs,W. Bernard,F. Horster
DOI: https://doi.org/10.1530/EJE.0.1310462
1994-11-01
European Journal of Endocrinology
Abstract:The objectives were to determine whether the serum thyroglobulin (TG) level is influenced by age or by non-thyroidal illness (NTI) of the aged, to investigate the constancy of the TG level after 1- and 2-month intervals and to investigate if the TG level could help to differentiate whether a subnormal thyrotrophin (TSH) level in a geriatric patient is caused by autonomous thyroid function, by age or by NTI. Two-hundred and twenty-six non-selected, chronic hospitalized patients over 60 years old and 82 healthy adults (20-40 years) participated in the study, and TSH, thyroxine, free thyroxine, triiodothyronine and TG were estimated. In 122 euthyroid geriatric patients with normal TSH the mean TG was normal (12.18 micrograms/l), but elevated (> 45 micrograms/l) TG values occurred more often than in healthy control persons (15/122 vs 3/82; chi 2(1) = 4.54, p = 0.03). The severity of the clinical state of the euthyroid patients had no influence on the TG values. If TG was measured after 1 and/or 2 months, in only 3/123 non-selected geriatric patients was there a fluctuation between the normal and abnormal range (versus fluctuation of the corresponding TSH values in 19/123 cases; chi 2(1) = 12.78, p = 0.0012). In 28 patients with subnormal TSH, a normal TG value had a predictive value of 0.6 to exclude autonomous thyroid function. Age and NTI of the geriatric patients have no significant influence on their mean TG level but high TG levels occur more often, even in euthyroid patients. The predictive value of TG is not sufficiently high to allow a clear differentiation of whether a subnormal TSH is caused by autonomous thyroid function or by the age process or by NTI. Nevertheless, the advantage of TG estimation to be more constant than TSH could be of benefit in screening studies.
What problem does this paper attempt to address?