Hormone replacement therapy with L-thyroxine promotes working memory and concentration in thyroidectomized female patients after differentiated thyroid carcinoma

Introduction

It is well established that long-lasting hypothyroidism is associated with cognitive deficits.^1–5 In their assessment of seniors with hypothyroidism, Osterweil et al. ⁴ found that patients showed significant impairments in learning, word fluency, visuo-spatial abilities, attention, visual scanning, and motor speed. Bjerke et al. ¹ demonstrated that impaired short-term memory in patients with a hypothyroid condition improved after L-thyroxine treatment. Aghili et al. ⁶ showed that L-thyroxine treatment in subclinical hypothyroidism led to improvement in cognitive function. In another important study, Kramer et al. ⁷ examined whether cognitive function and depressive symptoms differed between patients in who had normal thyroid function from those with hypothyroidism receiving long-term treatment with L-thyroxine and found no difference between the groups leading to the conclusion that hypothyroidism treated successfully improves cognitive functioning.

Based on these findings, we hypothesized that treatment with L-thyroxine in female patients with differentiated thyroid carcinoma would lead to improved cognitive performance compared to a hypothyroid state.

Method

Results

Metabolic status was determined using the TSH value. The mean TSH value in hypothyroidism (T1) was 55.52 mU/l (SD = 31.40), whereas in euthyroidism levels returned to normal (T2) 0.048 mU/l (SD = 0.119), Z = −3.296, p < 0.001, confirming that patients were in hypothyroid state during the first and in a euthyroid normal thyroid state during the second measurement point. Only one of the patients took oral contraceptives, three (21.4%) were in menopause, one (7.1%) was sterilized, three (21.4%) were in the luteal phase, two (14.3%) were in the follicular phase, and five (35.7%) were on their period.

A Wilcoxon signed-rank test revealed a significant difference in RCFT (cognitive reproduction) between the two groups and a tendency toward significance in D2 concentration performance and BDI (p = 0.046). No significant differences were found for other measures (see Table 1 for median, range, and test values).

OPEN IN VIEWER

Table 1.

Descriptive data and results of all comparisons.

	Hypothyroidism		Euthyroidism		Wilcoxon signed-rank test
	Median	Q25–Q75	Median	Q25–Q75	Z	p
RCFT (cognitive reproduction; points)	12.50	9.75–18.75	21.0	16.25–23.5	−3.183	0.001*
BDI (points)	4.50	1.00–7.25	1.50	0.75–5.25	−1.998	0.046
D2 concentration (points)	157.0	134.8–189.3	167.0	153.5–213.5	−1.992	0.046
Wechsler Digit Span (points)	13.0	10.0–16.25	14.5	13.0–17.0	−1.897	0.058
TMT-A (seconds)	31.58	22.63–36.28	24.27	19.62–34.65	−1.664	0.096
D2 total (points)	480.0	420.3–586.8	485.5	429.3–580.0	−1.444	0.149
VLMT (points)	49.0	43.25–57.0	53.0	49.0–58.25	−1.437	0.151
RWT II (points)	18.0	15.0–23.25	19.0	16.0–25.0	−0.946	0.344
TMT-B (seconds)	65.05	53.09–90.08	59.91	48.42–71.34	−0.847	0.397
RCFT (visual copying; points)	33.0	28.0–33.0	30.0	29.0–33.0	−0.631	0.528
RWT I (points)	21.0	14.25–26.0	21.0	19.75–24.5	−0.410	0.682

Note: Median, interquartile range (Q25–Q75), and test values of the comparisons between hypothyroidism and euthyroidism sorted in the order of significance. Significant at the α/N = 0.005 level. Significant p values are marked by asterisk (*). RCFT: Rey–Osterrieth complex figure test; BDI: Beck’s Depression Inventory; TMT: trail making test; VLMT: verbal learning and memory test; RWT: Regensburg word fluency test.

A Spearman correlation between BDI and RCFT cognitive reproduction was significant in hypothyroidism (ρ = 0.808, p < 0.001) and nonsignificant once thyroid levels returned to normal (ρ = 0.192, p = 0.530).

Moreover, a chi-square test showed no significant relationship between the difference between the two measurement points (as a measurement of change) of the significant changing psychometric measures and menstrual cycle, p ≥ 0.334 (data not shown), and between tumor stage, p ≥ 0.189 (data not shown).

Discussion

Altogether, we found that cognitive performance partly improved after substitution with L-thyroxine in thyroidectomized female patients after differentiated thyroid carcinoma compared to a hypothyroid state. In addition, we found an improvement in BDI scores. Importantly, however, depression levels were below clinically relevant values at both measuring time points excluding the possibility that our patients were depressed. ¹⁰ A comparison of correlations between the RCFT and BDI at both measurement points suggests a confounding of cognitive function by depressive symptoms in the hypothyroid state. However, unusually we found that higher BDI scores were associated with better cognitive performance. This may have been due to the small sample size. There is growing evidence of interactions between cognition and depression with many factors potentially contributing. ¹⁷ For example, Porter et al. ¹⁷ found cognitive impairment in young adult patients without antidepressant medication, which suggests a genuine relationship between depressive disorders and cognitive functioning. The hypothesized beneficial cognitive and neuropsychological effects of L-thyroxine substitution could only partially be confirmed. Beyond the Rey–Osterrieth complex figure test (cognitive reproduction) and a tendency toward significance in D2 concentration performance, the cognitive test battery showed no significant differences between the two thyroid states. The question remains why performance in other cognitive tests failed to show a significant improvement after return to normal thyroid levels. One possible explanation is that cognitive improvement is domain specific. In line with this idea, Schraml et al. ¹⁸ suggested that differences in cognitive performance exist between thyroidectomized patients in hypothyroidism and again following thyroid hormone replacement with regard to working memory. ¹⁹ They showed that the serum TSH level during hypothyroidism was inversely proportional to cognitive performance. Moreover, Samuels postulated that domains impaired by hypothyroidism are working memory and executive functions primarily. ²⁰ In functional magnetic resonance imaging studies deficits in working memory were no longer present after six months of substitution with L-thyroxine in patients with overt hypothyroidism ²¹ and subclinical hypothyroidism. ²²

Based on these results, it is possible that reduced working memory capacity was the underlying factor that influenced concentration and cognitive reproduction in our study. However, working memory was also an important prerequisite for many of the others tests, where we failed to find significant improvement.

A few methodological issues are important to note. Firstly, for the second measurement time point, despite the use of parallel versions, training effects might have distorted the results. Secondly, BDI scores correlated significantly with cognitive performance in hypothyroidism. Thus, learning effects and depressive mood are possible confounding factors of our results. Moreover, the use of a nonstructured psychiatric interview that was used to evaluate psychiatric diagnoses and medication intake could have led to a distortion of results, as could the small sample size limiting the generalizability of our findings. Thus, the study should be repeated with a larger sample and more standardized psychiatric interview procedures.

Based on our findings, we confirm that hormone replacement therapy with L-thyroxine promotes concentration and cognitive reproduction in thyroidectomized female patients after differentiated thyroid carcinoma. Clinical implications include routine cognitive and depression examination to improve quality of life and treatment of thyroidectomized patients.