Free Thyroxine, Brain Frailty and Clock Drawing Test Performance in Patients With Acute Minor Stroke or Transient Ischaemic Attack

Zhiliang Guo; Guoli Xu; Ruojun Wang; Jie Hou; Shuhong Yu; Huaishun Wang; Shuai Yu; Jiaping Xu; Shoujiang You; Zhichao Huang; Guodong Xiao; Yongjun Cao; Chun-Feng Liu

Disclosures

Clin Endocrinol. 2022;96(2):175-183.

Abstract and Introduction

Abstract

Objective: Thyroid dysfunction is associated with an elevated risk of cognitive decline, but the mechanism underlying this relationship is elusive. In this study, we investigate the relationships between free thyroxine (FT4), brain frailty and clock drawing test (CDT) performance in patients with acute minor stroke or transient ischaemic attack (TIA).

Design, Patients and Measurements: A total of 204 consecutive patients admitted to our hospital within 72 h after the onset of acute minor stroke or TIA were prospectively enroled and categorized in terms of quartiles of FT4 between March 2018 and August 2019. Brain frailty on magnetic resonance imaging was rated according to previously published criteria. Cognitive performance was assessed with the CDT.

Results: Generalized linear analysis revealed that FT4 was independently associated with higher brain frailty score after adjusting potential confounders (β, 0.03; 95% confidence interval [CI], 0.00–0.06; p = 0.0205), which is consistent with the result of FT4 (quartile) as a categorical variable (β, 0.34; 95% CI, 0.01–0.68; p = 0.0059; p _trend = 0.0807). A nonlinear relationship was detected between FT4 and brain frailty score, which had an inflection point of 1.19. FT4 was also associated with poor CDT performance (odds ratio, 1.15; 95% CI, 1.04–1.26; p = 0.0051). And mediation analysis found that brain frailty partially mediated the positive relationship between FT4 and poor CDT performance (indirect effect = 0.0024; 95% CI, 0.0003–0.01, p = 0.04).

Conclusions: Our findings suggested that a higher FT4 level was associated with a higher brain frailty score and poorer CDT performance, and brain frailty might play an important effect on the association between FT4 and cognitive decline.

Introduction

Thyroid dysfunction can cause a range of mood and cognitive disturbances, including anxiety, irritability, depression and deficits in executive function.^[1–3] Cognitive impairment and thyroid dysfunction are both associated with advancing age and are more prevalent in women.^[4,5] The evidence related to the association between thyroid dysfunction and cognitive impairment is conflicting. A previous study showed hypothyroidism was associated with an increased risk of dementia.^[6] However, there were studies that revealed there was a significant association between hyperthyroidism and cognitive decline.^[5] The systematic review and meta-analysis indicated that hyperthyroidism, but not hypothyroidism, might be associated with a modestly elevated risk of dementia.^[2] In particular, the elevated free thyroxine (FT4) has been shown to be associated with an increased risk of cognitive disturbances.^[5,7] The underlying mechanisms explaining the link between thyroid function and cognitive disturbances are yet unclear, but possible and yet unexplored pathways are through subclinical changes in microstructural organization or brain tissue atrophy.^[7]

Brain frailty is usually conceptualized as the product of physiologic changes associated with advancing age and the accumulation of multimorbidity of underlying brain pathology in patients presenting within the typical stroke age bracket.^[8,9] This causes a loss of resilience to acute health problems that may result from both diseases and extrinsic stressors. Recently, a 3-point score that included 1 point for white matter hyperintensities (WMH), 1 point for atrophy and 1 point for any chronic infarct or vascular lesion was created by Appleton et al. to define brain frailty.^[10] And the increased baseline brain frailty score is associated with worse cognitive outcomes at 90 days in acute stroke participants.^[10] In addition, there is evidence that thyroid dysfunction is associated with increased cerebral small vessel disease (CSVD), including WMH and lacunes, which is likely to be one factor contributing to brain frailty.^[11,12] In this view, it is plausible to assume that an increased brain frailty score may be in the path of the association between elevated FT4 and cognitive decline. However, such a relationship has not been fully understood.

Using data from a cohort of acute minor stroke or transient ischaemic attack (TIA) population, we aimed to assess the effect of a high brain frailty score on the association between FT4 and the clock drawing test (CDT) performance in stroke patients. The CDT was chosen as the dependent variable because it provides an economical and comprehensive evaluation of multiple cognitive domains.^[13]

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Baseline characteristics of study participants according to quartiles of FT4
Characteristics	Q1	Q2	Q3	Q4	p Value
No. of patients	125	135	130	131
Age, years, mean (SD)	68.61 ± 11.12	65.02 ± 10.81	64.00 ± 10.42	61.65 ± 11.52	0.015
Female (%)	16 (31.37%)	22 (43.14%)	22 (43.14%)	18 (35.29%)	0.524
BMI (kg/m²), mean (SD)	25.21 ± 4.28	24.91 ± 3.54	24.63 ± 3.47	24.47 ± 3.19	0.749
Education status					0.579
Illiterate	21 (41.18%)	28 (54.90%)	24 (47.06%)	20 (39.22%)
Primary school	16 (31.37%)	10 (19.61%)	12 (23.53%)	19 (37.25%)
Junior high school	10 (19.61%)	6 (11.76%)	10 (19.61%)	9 (17.65%)
High school	4 (7.84%)	6 (11.76%)	5 (9.80%)	3 (5.88%)
University degree or above	0 (0.00%)	1 (1.96%)	0 (0.00%)	0 (0.00%)
Hypertension (%)	46 (90.20%)	39 (76.47%)	32 (62.75%)	41 (80.39%)	0.010
Diabetes (%)	13 (25.49%)	13 (25.49%)	11 (21.57%)	14 (27.45%)	0.920
Hyperlipidemia (%)	23 (45.10%)	33 (64.71%)	33 (64.71%)	33 (64.71%)	0.106
Coronary heart disease (%)	0 (0.00%)	2 (3.92%)	0 (0.00%)	1 (1.96%)	0.293
Atrial fibrillation (%)	5 (9.80%)	3 (5.88%)	1 (1.96%)	1 (1.96%)	0.201
Smoking (%)	25 (49.02%)	22 (43.14%)	25 (49.02%)	26 (50.98%)	0.872
Drinking	9 (17.65%)	3 (5.88%)	6 (11.76%)	11 (21.57%)	0.116
Baseline NIHSS, median (IQR)	2.00 (0.00–2.00)	1.00 (0.50–2.00)	1.00 (0.00–2.00)	2.00 (1.00–2.00)	0.469
Intracranial arterial stenosis (%)	17 (33.33%)	20 (39.22%)	17 (33.33%)	17 (33.33%)	0.900
Extracranial arterial stenosis (%)	12 (23.53%)	9 (17.65%)	5 (9.80%)	5 (9.80%)	0.152
IV rtPA (%)	3 (5.88%)	3 (5.88%)	6 (11.76%)	2 (3.92%)	0.430
eGFR (ml/min/1.73 m²), median (IQR)	74.85 (61.39–103.72)	83.57 (73.82–103.59)	86.10 (74.88–108.84)	103.24 (71.76–113.04)	0.049
Homocysteine (μmol/L), median (IQR)	12.20 (9.75–15.15)	9.20 (7.55–11.95)	10.10 (8.40–12.10)	10.00 (8.25–13.95)	0.002

Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate; FT4, thyroxine; IQR, interquartile range; IV rtPA indicates intravenous recombinant tissue plasminogen activator; NIHSS, National Institutes of Health Stroke Scale; SD, standard deviation.

Table 2. Relationship between FT4 and brain frailty among patients with acute minor stroke or transient ischaemic attack in different models
Variable	Nonadjusted model		Model 1		Model 2
Variable	β (95% CI)	p Value	β (95% CI)	p Value	β (95% CI)	p Value
FT4	0.00 (−0.02, 0.03)	0. 7632	0.03 (0.00, 0.05)	0.0333	0.03 (0.00, 0.06)	0.0205
FT4 (quartile)
Q1	0		0		0
Q2	0.20 (-0.18, 0.57)	0.3106	0.38 (0.06, 0.71)	0.0223	0.48 (0.14, 0.82)	0.0059
Q3	0.14 (-0.24, 0.52)	0.4775	0.37 (0.04, 0.70)	0.0278	0.48 (0.14, 0.82)	0.0064
Q4	0.00 (-0.38, 0.38)	1.0000	0.32 (-0.01, 0.65)	0.0586	0.34 (0.01, 0.68)	0.0479
p _trend	0.9232		0.0815		0.0807

Note: Nonadjusted model: we did not adjust other covariates. Model 1: we adjusted age and sex. Model 2: we adjusted age, sex, BMI, education status, hypertension, diabetes, hyperlipidemia, baseline NIHSS, intracranial arterial stenosis, extracranial arterial stenosis, eGFR and homocysteine.
Abbreviations: BMI, body mass index; CI, confidence interval; eGFR, estimated glomerular filtration rate; FT4, thyroxine; NIHSS, National Institutes of Health Stroke Scale.

Table 3. The results of the two-piecewise linear regression model
Outcome	Effect size, β (95% CI)	p Value
Inflection point of FT4	0.28
< 1.19	1.83 (0.16, 3.50)	0. 0326
≥1.19	−0.01 (−0.06, 0.03)	0.5989
P for log likelihood ratio test	0.024

Note: Effect: FT4; Cause: brain frailty score; Adjusted: age, sex, BMI, education status, hypertension, diabetes, hyperlipidemia, baseline NIHSS, intracranial arterial stenosis, extracranial arterial stenosis, eGFR and homocysteine.
Abbreviations: BMI, body mass index; CI, confidence interval; eGFR, estimated glomerular filtration rate; FT4, thyroxine; NIHSS, National Institutes of Health Stroke Scale.

Table 4. Relationship between FT4 and the CDT performance among patients with acute minor stroke or transient ischaemic attack in different models
Variable	Non-adjusted model		Model 1		Model 2
Variable	β/OR (95% CI)	p Value	β/OR (95% CI)	p value	β/OR (95% CI)	p Value
CDT score
FT4	−0.04 (−0.07, 0.00)	0.0789	−0.04 (−0.08, 0.00)	0.0518	−0.05 (−0.08, −0.01)	0.0206
FT4 (quartile)
Q1	0		0		0
Q2	−0.31 (−0.84, 0.21)	0.2453	−0.29 (−0.81, 0.22)	0.2662	−0.23 (−0.74, 0.28)	0.3819
Q3	−0.49 (−1.02, 0.04)	0.0702	−0.49 (−1.00, 0.03)	0.0662	−0.59 (−1.11, −0.07)	0.0263
Q4	−0.39 (−0.92, 0.14)	0.1468	−0.48 (−1.00, 0.04)	0.0706	−0.46 (−0.97, 0.05)	0.0774
p _trend	0.1129		0.0518		0.0376
Poor CDT performance
FT4	1.09 (1.02, 1.17)	0.0176	1.11 (1.02, 1.20)	0.0101	1.15 (1.04, 1.26)	0.0051
FT4 (quartile)
Q1	1.0		1.0		1.0
Q2	2.62 (0.96, 7.12)	0.0590	2.63 (0.94, 7.38)	0.0658	2.16 (0.64, 7.27)	0.2150
Q3	3.43 (1.28, 9.16)	0.0140	3.67 (1.32, 10.19)	0.0126	5.33 (1.56, 18.19)	0.0075
Q4	2.62 (0.96, 7.12)	0.0590	3.20 (1.12, 9.14)	0.0294	3.81 (1.10, 13.12)	0.0342
p _trend	0.0586		0.0258		0.0155

Note: Nonadjusted model: we did not adjust other covariates. Model 1: we adjusted age and sex. Model 2: we adjusted age, sex, BMI, education status, hypertension, diabetes, hyperlipidemia, baseline NIHSS, intracranial arterial stenosis, extracranial arterial stenosis, eGFR and homocysteine. Poor CDT performance was defined as CDT score ≤ 2.
Abbreviations: BMI, body mass index; CDT, clock drawing test; CI, confidence interval; eGFR, estimated glomerular filtration rate; FT4, thyroxine; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio.