| Objective Thyroid function abnormality is a clinically common disease, and includes hypothyroidism, subclinical hypothyroidism, hyperthyroidism, and subclinical hyperthyroidism. With the bigger pressure and the faster pace in modern life, the incidence of thyroid dysfunction is obviously on the rise. The prevalence of thyroid function abnormality in women populations is higher than in men. Many researcher focused on the relationship between thyroid dysfunction and cognition. Previous literature demonstrated the effect of thyroid hormones on the brain function from study of behavior, neuroelectrophysiology and brain imaging. However, it is controversial whether there is an association between thyroid dysfunction and cognitive impairments. The purpose of our study is to determine whether cognitive functions are impaired in patients of thyroid function abnormality using attention networks test(ANT) and electrophysiological technique(Event-related Potentials, ERP).Methods Study 1. 31 patients with hypothyroidism, 32 patients with subclinical hypothyroidism, 34 patients with hyperthyroidism, 30 patients with subclinical hyperthyroidism and 42 matched healthy controls with euthyroidism performed the attention network test(ANT), which was used to assessed attentional ability. Study 2. 18 patients with newly diagnosed hypothyroidism are treated with levothyroxine sodium, and 25 patients with newly diagnosed hyperthyroidism are treated with methimazole. We compared the performance of ANT in patients with hypothyroidism and with hyperthyroidism before and after treatment. We also selected 31 matched healthy controls with euthyroidism. Study 3. 25 patients with hyperthyroidism and 20 matched healthy controls with euthyroidism performed signal stop task(SST) and gambling task(GT). We used the ERP technique to explore the electrophysiological mechanism of the effect of hyperthyroidism on cognition.Results Study 1. There was significant differences for alerting network between 5 groups(p<0.05). The mean scores for alerting network was significantly lower in patients with hypothyroidism and hyperthyroidism than in healthy controls(p=0.007; p=0.001). The mean reaction time for executive control network was significantly longer in hyperthyroid patients than in healthy controls(p=0.006). There were no significant differences in the mean score of orienting network, overall mean reaction time and accuracy between 5 groups(p>0.05). Triiodothyronine(T3) and thyroxine(T4) were positively correlated with the mean scores for executive control network in all the subjects(r=0.22, p=0.004; r=0.21, p=0.007). T4 was negatively correlated with the mean scores for executive control network in patients with hypothyroidism(r=-0.47, p=0.008). T4 was positively correlated with the mean scores for executive control network in patients with hyperthyroidism(r=0.36, p=0.038). Study 2. There were significant differences for alerting networks between patients with hypothyroidism before treatment and healthy controls(p<0.05), and there were no significant differences in the mean score of orienting and executive control networks, overall mean reaction time and accuracy between 2 groups(p>0.05). The mean score of alert network was significantly higher in hypothyroid patients after treatment than before treatment(p<0.05). There were no significant differences in the mean score of orienting and executive control networks, overall mean reaction time and accuracy in patients with hypothyroidism before and after treatment(p>0.05). There were significant differences for alerting and executive control networks between patients with hyperthyroidism before treatment and healthy controls(p<0.05; p<0.01) and there were no significant differences in the mean score of orienting network, overall mean reaction time and accuracy between 2 groups(p>0.05). The mean reaction time for executive control network was significantly shorter in hyperthyroid patients after treatment than before treatment(p<0.05). There were no significant differences in the mean score of alerting and orienting networks, overall mean reaction time and accuracy in patients with hyperthyroidism before and after treatment(p>0.05). Levels of T4 was negatively correlated with the scores for executive control network in patients with hypothyroidism before treatment(r=-0.473, p=0.047). Levels of T3 and T4 were positively correlated with the scores for executive control network in patients with hyperthyroidism before treatment(r=0.49, p=0.013; r=0.49, p=0.013). Level of TSH was negatively correlated with the score of HAMD in patients with hyperthyroidism(r=-0.53, p=0.006). Study 3. In signal stop task, there was no significant difference for mean probability of successful inhibitions on stop-trials between 2 groups(p>0.05). There were significant differences for mean percentage of correct responses on go-trials and of missed responses on go-trials between 2 groups(p<0.05). In gambling task, There was significant main effect for the blocks factor(F(5, 215)=4.44, p=0.001). There was no significantly in interaction effects for blocks factor and group(p>0.05). ERP results demonstrated there were significant main effect for stimuli type(F(1, 43)=20.56, p<0.001) and electrodes position(F(8, 344)=26.35, p<0.001), and interaction effects for stimuli type and group(F(1, 43)=5.05, p<0.05) in N2 amplitude. In P2 amplitude, there were significant main effect for stimuli type(F(1, 43)=20.52, p<0.001) and electrodes position(F(8, 344)=13.78, p<0.001),and interaction effects for stimuli type and group(F(1, 43)=12.45, p<0.01). In P3 amplitude, there were significant main effect for stimuli type(F(1, 43)=205.47, p<0.001) and electrodes position(F(8, 344)=18.98, p<0.001). There was on significant interaction effect for stimuli type and group(p<0.05). In FRN component,there were significant main effect for feed type(F(1, 43)=27.84, p<0.001), intensity(F(1, 43)=35.45, p<0.001) and electrodes position(F(2, 86)=63.77, p<0.001). There were significant interaction effects for feed type, intensity and group(F(1,43)=4.45, p<0.05). In P3 component, there were significant main effect for feed type(F(1, 43)=47.93, P<0.001), intensity(F(1, 43)=27.95, p<0.001) and group(F(8, 344)=45.55, p<0.001). There was no significant interaction effect for feed type, intensity and electrodes position(p>0.05).Conclusions Study 1. The patients with hypothyroidism existed alerting network impairments. The patients with hyperthyroidism existed alerting and executive control networks impairments. We did not find attention deficits in subclinical hypothyroidism and subclinical hyperthyroidism. There was an association between thyroid hormones and executive control function. Study 2. The patients with untreated with hypothyroidism had alerting networks impairment, and the impairment improved after treatment with levothyroxine sodium. We found the patients with untreated hyperthyroidism had alerting and executive control networks impairments. The executive control function might improve after treatment with methimazole. However, there was no significant difference for the score of alerting network in patients with hyperthyroidism before and after treatment. Study 3. The current study demonstrated the performance in executive function of patients with hyperthyroidism were worse than of healthy controls. N2 amplitude was larger in healthy controls than in patients with hyperthyroidism in signal stop task. The sensitivity to feedback intensity had reduced in hyperthyroism in gambling task. |
PDF Full Text Request