We investigated the natural course of subclinical thyroid dysfunctions in geriatric patients, especially regarding their association with mortality rate. Ninety-three randomly selected chronically ill geriatric patients 6487 (median: 77) yr of age participated in the screening study with a 2-yr follow-up. Serum thyrotropin (thyroid-stimulating hormone [TSH]), free thyroxine, triiodothyronine, and antibodies against thyroid peroxidase were measured. During the follow-up, patients with suppressed TSH levels who were otherwise euthyroid (untreated) had a higher mortality rate than patients with normal TSH (5/8 vs 18/64; p < 0.05). The initial clinical state of these two subgroups did not differ significantly. Two-thirds of patients with treated hyperthyroidism died. The mortality rate of patients with initially subnormal but not suppressed TSH level was average and did not differ statistically from either the euthyroid or the hyperthyroid groups. Only 1 of 13 euthyroid patients with positive thyroid antibody titers developed a subsequent subclinical hypothyroidism. Subclinical hyperthyroidism was found to be associated with a higher mortality rate in chronically ill geriatric patients, which justifies screening for thyroid dysfunction and treatment of subclinical hyperthyroidism. in addition, a subnormal but measurable TSH was not indicative regarding the future development of hyperthyroidism. Finally, during the 2-yr follow-up, antibody positivity in the euthyroid cases did not prove to be predictive for the subsequent development of hypothyroidism.

UNLABELLED The objective of this study was to investigate if screening of chronically ill geriatric patients for thyroid dysfunction is justified just upon hospital admission. TSH was measured in 124 patients at hospital admission and 11-86 (Median 37) days afterwards. FT4 was measured in cases with subnormal, suppressed or elevated TSH (43 cases). Out of 81 patients with normal (0.5-3.6 mU/l) TSH, the control value was subnormal (0.1-<0.5 mU/l) in 6 and elevated (>3.6 mU/l) in one case, but in none of the patients became suppressed (<0.1 mU/l). In 13/30 patients with subnormal TSH the control value was normal but in none of the patients suppressed or elevated. On the contrary, all cases with suppressed (N=9) or elevated (N=4) TSH remained in the same ranges at follow up. Low (<13 pmol/l, N=3) or elevated (>27 pmol/l, N=5) initial FT4 levels did not change in the follow up as well. Out of 35 patients with normal FT4, one became low and another elevated. Improvement or worsening of the clinical state in the follow up did not correlate to changes of TSH. The prevalence of unsuspected thyroid dysfunctions were 11.3% (hyperthyroidism clinical: 4, subclinical: 5, hypothyroidism clinical: 3, subclinical: 2 cases). All cases except one with subclinical hypothyroidism were detected by the initial screening. Only one patient with clinical hyperthyroidism was initial misinterpreted as having subclinical disease. CONCLUSIONS In chronically ill geriatric patients investigated at hospital admission, a measurable TSH practically excludes hyperthyroidism in the follow up. Suppressed TSH levels remain suppressed but subnormal levels should be controlled because their normalization frequently occur in the follow up. Screening upon hospital admission is sensitive enough to detect cases of thyroid dysfunction and justified by their high prevalence.

OBJECTIVE: To investigate the effect of age, clinical condition, and thyroid function on the prevalence of thyroid autoantibody positivity in hospitalized chronic geriatric patients. DESIGN: A screening study of hospitalized chronic geriatric patients. PARTICIPANTS: 249 non-selected, hospitalized, chronic geriatric patients more than 60 years of age and 81 20 to 40-year-old healthy persons. MEASUREMENTS: Thyrotropin (TSH); thyroxine (T-4) and free thyroxine (FT4); and triiodothyronine (T-3), thyroglobulin (Tg), antibodies against thyroid peroxidase (AbTPO) and antibodies against thyroglobulin (AbTg) estimation in a screening study.RESULTS: AbTPO positivity (AbTPO+) was found more often than AbTg positivity (AbTg+) (15.3% vs 9.2%, P = .04), one being positive (Ab+) in 19.3%. The occurrence was higher in females than males (Ab+:27.1% vs 7.1%, P < .001; AbTPO+:21.9% vs 5.1%, P < .001; AbTg+:13.2% vs 3.1%, P = .0052). Among the Ab+ patients, AbTPO was more often positive than AbTg (40/48 vs 21/48, P (.001). The sensitivity, specificity, and positive predictive value of Ab positivity to detect a thyroid disorder were 0.35, 0.85 and 0.38, respectively. Within the population of euthyroid geriatric patients, the occurrence of AbTg+ (chi((2)(2)) = 8.65, P = .013) and Ab+ (chi((2)(2)) = 8.02, P = .018) correlated positively with the age of the patients, and there was also a female predominance (AbTPO+ 18% vs 3.7% in the males; AbTg+ 13% vs 2.4%; Ab+ 25.8% vs 6.1%). When compared with 20 to 40-year-old subjects, only the euthyroid greater than or equal to 80-year-old patients showed a significantly higher occurrence of Ab+ (26.2% vs 9.9% chi((1)(2)) = 5.64, P = .017). In the euthyroid greater than or equal to 80-year-old females, AbTPO+ was 25%, AbTg:22.2%, and Ab+: 36.1%!. The nonthyroidal clinical state of the euthyroid patients did not correlate with the antibody prevalence.CONCLUSIONS: In hospitalized chronic geriatric patients, AbTg and especially AbTPO positivity is frequent, even in euthyroid patients without goiter. This aspecific Ab positivity in the euthyroid state correlates to the age, but not to the severity of the nonthyroidal clinical condition of the patients and explains why the Ab positivity is not predictive enough for thyroid dysfunction in this subpopulation. Thus, in hospitalized chronic geriatric patients the AbTg and AbTPO titers should be examined only in cases where thyroid screening (TSH) reveals abnormal results.

Szabolcs I, Bernard W, Horster FA. Serum thyroglobulin in hospitalized chronic geriatric patients: its relationship to age, non-thyroidal illness, goitre and thyroid dysfunction in a follow-up study. Eur J Endocrinol 1994;131:462–6. ISSN 0804–4643 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 μg/l), but elevated (> 45 μg/l) TG values occurred more often than in healthy control persons (15/122 vs 3/82; χ (1) 2 = 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; χ (1) 2 = 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. I Szabolcs, HIETE I. Bel., Budapest, PO Box 112, 1389, Hungary