Endocrine Abstracts

Background: Meta-analyses have questioned whether mortality is increased in hyper- and hypothyroid individuals, and if so the magnitude. This is mainly due to inadequate size of study populations, differences in study design and participant characteristics, lack of control for confounders and whether the participants are treated.

Materials: Using record-linkage data from nationwide Danish health registers, and after adjustment for preexisting morbidity, a 22–42% excess mortality was demonstrated in hyperthyroidism, higher in Graves’ disease than in toxic nodular goiter. In hypothyroidism the excess mortality was 23%. In both phenotypes there was an increased risk of being diagnosed with a number of morbidities, especially cardiovascular diseases. For both hyper- and hypothyroid individuals there was an increased risk - 51 and 140%, respectively - of being diagnosed with psychiatric diseases (psychosis, depression and anxiety). Both hyper- and hypothyroidism was associated with an increased risk - 88 and 89%, respectively - of receiving disability pension. The above studies lacked adequate thyroid function variables and data on therapy. In a register-based Danish cohort of ca. 235.000 individuals offering such data, and followed for a median of 7 yr., increased mortality (23%) in untreated but not in treated hyperthyroid patients was demonstrated. Hazard ratio for mortality was around 1.12 for every 6 months of decreased TSH, whether individuals were treated or not. In the same cohort, untreated hypothyroid individuals had an excess mortality of 46%, which was at large uninfluenced by age (below or above 65 yr.) and severity (mild or overt) of hypothyroidism. Mortality was increased by 5% for every 6 months of increased TSH.

Conclusions: Mortality is increased in both hyper- and hypothyroidism. Cumulative periods of abnormal TSH increases mortality in both treated and untreated individuals, suggesting that not lack of therapy but lack of maintaining euthyroidism may drive excess mortality.