Brain, mood and cognition during treatment initiation in mild hypothyroidism

Description: This collection is published here: Endocrine Abstracts (2015) 37 EP1049. It describes an additional study group from the study described here, Quinque et al. 2014: doi 10.1016/j.psyneuen.2014.01.015 Patients were newly diagnosed with hypothyroidism due to Hashimoto's thyroiditis. 11 patients were recruited between June 2011 and June 2013 via local endocrinologists and among blood donors and participants of a large cohort study (LIFE). One patient had to be excluded due to a benign brain anomaly, another for a lack of hypothyroidism at study entry. Nine healthy control subjects were selected from the control sample of Quinque et al. 2014 to match the patients in age, sex and intelligence. Sample description: mean age/SD patients 42+/-5, controls 38+/-8; 8 females each; IQ patients 104+/-10, controls 107+/-8. Thyroid laboratory values: patients session 1: TSH 6.1+/-1, fT3 4.8+/-1, fT4 12.6+/-2, TPO 285+/- 184; patients session 2: TSH 3.4+/-1, fT3 4.6+/-1, fT4 15.7+/-2, TPO 287+/-184; controls session 1: TSH 2.9+/- 1, fT3 4.5+/-1, fT4 14.9+/- 1, TPO 11+/-2; controls session 2: TSH 2.5+/- 1, fT3 4.4+/-0.4, fT4 14.5+/- 1, TPO 9+/-3. The scanning protocol was identical to the one reported in Quinque et al. 2014. Analysis was identical except that data was not averaged over the two timepoints. Groups were compared at both sessions separately. In addition, paired t-tests were used to analyse changes over time in the patient group. Thirdly, the interaction between session and group was analysed in a 2x2 ANOVA design. All analyses were controlled for age and sex, VBM analyses additionally for total intracranial volume (TIV). Resting-state fMRI was only available for 8 patients, because one patient panicked after the structural scan. One control subjects was equally removed from the analysis to ensure matching between the groups. The task fMRI data was not analysed due to an unexpected performance difference even before treatment in the direction that the patients performed better. Overall conclusion: We were not able to find any group differences or changes over time that would survive classical statistical thresholds at the level of whole brain analyses or regions of interest analyses, (six ROIS as in Quinque et al. 2014 were used). As the sample was small and the disease severity and treatment effect on thyroid laboratory values likewise, we cannot conclude whether the result is due to a lack of power or whether there really are no structural and functional brain changes. Therefore, we refrain from publication, but want to make the data available for metaanalyses to avoid publication bias towards positive findings from little powered studies.

Related article: http://doi.org/10.1530/endoabs.37.EP1049