Changes of Visual Evoked Potentials in Patients with Thyroid Dysfunction

CITTO IULIAN TAISESCU1*, VIOREL BICIUSCA2, LORENA SAS3, GEORGIANA CAMEN4, TEODOR SAS4, CARMEN ALBU5, OANA TAISESCU3 1University of Medicine and Pharmacy of Craiova, Department of Physiology, 2 Petru Rares, 200349,Craiova, Romania 2University of Medicine and Pharmacy of Craiova, Department of Semiology, 2 Petru Rares, 200349, Craiova, Romania 3University of Medicine and Pharmacy of Craiova, Department of Anatomy, 2 Petru Rares, 200349, Craiova, Romania 4University of Medicine and Pharmacy of Craiova, Department of Radiology, 2 Petru Rares, 200349, Craiova, Romania 5University of Medicine and Pharmacy of Craiova,Department of Neurology, 2 Petru Rares, 200349, Craiova, Romania

Evoked potentials can be defined as voltage variations that can be recorded at the level of some components of the nervous system in response to different sensory stimulation modalities [2][3][4]. The visual evoked potentials technique records the cerebral bioelectric activity generated by light stimulation [5,[7][8][9]. Visual evoked potentials is a electrophysiological test [10,11], a good method of assessing the electrical response of the brain to visual stimuli for evaluation of impulse travel from the eye to the occipital cortex [40].
Thyroid hormone are involved in myelin production, axonal transportation, and neurotransmission [13][14][15]. Variations in thyroid hormone concentrations may cause abnormalities in the perception and conduction of impulses from the receptor level to the occipital cortex [35].
We studied the visual evoked potentials in 18 patients with hyperthyroidism and 18 patients with hypothyroidism, also, we compared hyperthyroid patients with hypothyroid patients [17,18] and we studied not only the P100 wave and the N75 wave latencies, but also all of their 7 parameters.

Experimental part Materials and methods
This study was conducted in the period 2014-2017, through the collaboration between the Department of Physiology of UMF Craiova and the Department of Internal Medicine of Emergency Military Clinic Hospital Stefan Odobleja, from Craiova. Thirty-six patients with thyroid disease were investigated by the visual evoked potential, which of 18 patients with hypothyroidism and 18 patients with hyperthyroidism. The subjects presented an obvious thyroid pathology, diagnosed based on clinical manifestations and paraclinical determinations [6,19,20]. All patients had evidence of medium-intensity thyroid disease without complications of cardiovascular, ophthalmological, psychiatric or hematological nature [12,16]. It is known that antithyroid drugs have been associated with development of agranulocytosis or secondary autoimmune neutropenia [1]. The questionable cases, with early-stage disease with unsafe diagnostics, with manifestations at the limit, have been removed. Records were made on each subject once, as soon as possible after hospital admission, 1-2 days after treatment commenced. The recording conditions have always been the same of thermal comfort, semi-security, lack of noise and mental relaxation [22,23].
The modifications of visual evoked potentials (VEP) were determined by recordings of these potentials in parallel with the hormonal, biochemical and physiological manifestations (pressure, reflexes) [25,26], on a total number of 36 patients with thyroid pathology, 18 with hyperthyroid and 18 with hypothyroid. The male gender predominated with 27 cases (75%). The predominance of the male gender was even higher on the hyperthyroid lot (16 cases, that is 88.89%), being lower at hypothyroid (11 cases-61.11%) (table 1).
The sanguine concentration of the T3 and T4 hormones were dosed using classic methods [28,33]. Lipemia and cholesterolemia were determined with the help of classical biochemical analysis, on blood collected a jeune, under identical conditions for all subjects and at the same time of the day. The basal metabolism was determined under rigorously kept basic conditions, through a modern device [30,39]. The Achilean reflexogram was recorded with an efficient dedicated electronic device, and by classic clinical methods the cardiac frequency and the minimum and maximum blood pressure were measured, in a state of basal rest, at the same time of the day.
The subject is in absolute physical and psychical rest, in a dusky room, free of noise, in thermal balance, with no hunger, third, warm-cold, pain (head ache, arthralgia) sensations, with no thyroid medication for 6 days, with no exciting, depressing substances (coffee etc.), tranquillizers, neurotropic vitamins [32,36], the subject sitting comfortably on a chair and looking attentively at the screen through which the repeated photic stimuli are applied. If the subject has refraction vices, he will look through the vice-correcting glasses.
Placing of the EEG electrodes was performed according to figure 1, after degreasing the skin and applying electroconductive paste, the electrodes being fixed by collodion gluing ( fig.1). We used a Nihon-Kohden Neuropack M1 with Neurofax module.
In order to extract VEP signals from the EEGs, we recorded 300 ms after applying the photic excitation. The sampling frequency was 1000 Hz (therefore one sample/ ms), and the analog-digital conversion (A/D) was done on 12 bits.
Normally 100-180 answers were summed. In order to precisely identify the level of the isoelectric line, we started adding the samples with 60 ms before the moment of applying the photic excitation technique. Therefore at each addition 360 samples were added, 60 before, 300 after the stimulation.
Of all the different techniques used for photic stimulation (flash, rotating mirror projector, TV screen, LED screen etc.), the on screen LED stimulation is the best, precise and physiological (2 and 5). The pattern reversal stimulation method was used, with the help of a matrix of approximately 10x15 cm, made from a few thousand yellow-green LEDs,. The photic stimulation frequency was of 1 Hz the highest.

Results and discutions
The average ages of the two groups were a little different, 40.89 +/-6.02 in hyperthyroid, with limits between 30 and 50 years and 38.22 +/-11.08 in hypothyroids, with limits between 22 and 50 years (table 2).
The age of the illness in studied hyperthyroid was 2.33 years +/-1.97, and the duration of the treatment 1.22 +/-1.77 years, and in hypothyroid the age was 6.06 +/-3.10 years and the duration of the treatment 5.36 +/-3.6 years (table 2).  The average ages of the two groups (with very close values) was not statistically significant, (Student t test p>0.05),while statistic differences between the average age of the illness and the average duration of the treatment were highly significant (p = 0.0005, and 0.0004 respectively) (table 3).
Obtained hormonal, iodine uptake, biochemical, metabolism, reflectivity and cardiovascular values. Tables 4, 5, 6, 7, 8 present the average values, standard deviations, the coefficient of variation % (CV%) and other current statistic parameters for all the performed determinations, the absolute and relative differences (%) between the averages of the two studies groups and the statistic insurances of these differences, and in figure 2, 3, and 4 the important data in the quoted tables is exposed graphically in absolute values then percents [21,27,31], hypothyroid as opposed to hyperthyroid considered comparison standard (100%).
In tables 4 and 5 great differences can be observed (complying with familiar clinical and paraclinical semiology) between the average values of hyper and hypothyroids.
Then it can also be observed that the majority of obtained mediums have a very high safety and informative value, CV% being below 10% (see yellow colored boxes), and the ones with CV% over 10% have 11, 15, 16, maximum 21%,  so not very high. Thus all the obtained mediums, generally speaking, are safe from an informative statistic point of view [24,29]. The differences between the average values obtained on the lot of hyperthyroid patients compared to the ones obtained on the lot of hypothyroid patients [37,41] are very high, between 32.02% and 140.85% -the majority are 40 ± 91% at nine comparisons, only at 2 they are lower than 13.91% and 6.44 (table 6).
There are significant statistical ensurances between the two lots for all the parameters.
These great differences are statistically significant [34,38], with very high significance (p<0.001) in most cases, as shown by 3 calculus methods (Student, ANOVA, Chi 2 ),   In figure 2, 3 and 4 the differences of the average absolute values between the 2 groups ( fig. 2, 3 ) and relative values % ( fig.4) can be seen.
Following these measurements, there was also a great mathematical correlation coefficient r between the majority (9) of studied parameters [42,45] (except for blood pressures), with values over 0.9 (0.902 to 0.986), correlations calculated by the Pearson formula [43,44]. That shows that if a parameter increases with 100%, the other one for which the correlation was calculated to the first one modifies (plus or minus) with 90.2% respectively 98.61%, which is an interesting scientific fact (table 9).
This precise mathematical determination of the thyroid function on the 2 groups was necessary to know exactly what thyroid pathology, how severe, we explore through VEP. Tables 10 and 11 show that in hypothyroid patients the latency of the P100 increases by 17.88%(p<0.01), and the one of the N75 wave only with 9.88% (p<0.05), as proved by the Student's t test, but also by the Mann-Witney and ANOVA tests (table 10).

Modifications of visual evoked potentials
Also, the duration of the N75 wave drops with 26.37% (p = 0.0265). All these differences of the VEP waves, statistically significant or not, are presented % in figure 5 In addition to the quoted authors we found important percentage differences, highly significant (p<0.001) for P100 and only significant (p<0.05) for N75, because we had greater differences of the thyroid function between hyper and hypothyroid patients, while all previously quoted article showed smaller differences between normal thyroid and hypothyroid patients (in most articles) or between normal thyroid and hyperthyroid patients. Moreover, some authors note statistically significant differences, others don't, but no information is given on the severity of thyroid pathology. Possibly the pathology was more severe for some authors, and easier for others, where the increase of the latency of the P100 wave was not statistically significant.
The other parameters of the VEP, other than N75 and P100 waves, although presenting great percentage differences, up to 52.35%, due to the high dispersion of the individual values obtained on them (with CV% between 25.17 and 151.16%, the majority around 80%), showed no statistic significant differences -as observed more clearly in table 11 and figure 5.
In order to interpret all the differences of the average values of all VEP parameters on the two main waves N75 and P100, even if they are not statistically significant (beside latency differences), we must start, in order to interpret  Student's t test, respectively how large the statistical ensurance of differences between the 2 patient lots is them, from the following safe scientific hypothesis: wave N75 is a VEP wave that expresses the excitation of the primary visual cortex generated by the sumation of the synaptic potentials, depolarized (negative), excitations generated by the synapses of the visual ways on the cortical neurons in field 17, while the VEP P100 wave appears by summing up hyper-polarizing inhibitory synapses (positive) produced by inhibitory synapses that come from satellite, inhibitory, cortical neurons, which, through their action, restrain chaotic, generalized spread of the excitation, limiting it only to the area on which it must command it, in which precise cortical processing of visual information are produced.
Latencies increase because, in hypothyroid patients, all neuronal processes, including excitation (especially through synapses), are delayed.
As soon as the latency of the N75 wave increases, the latency of the P100 wave, that is triggered by the N75 wave, forcibly increases higher and better statistically ensured, because the synaptic transmission between excited neurons (wave N75) and inhibitory satellite ones, as well as the transmission between recurrent fibers of inhibitory satellite neurons that come back on exciting neurons are also extended in time. It is precisely explained that the latency of the N75 wave increases only by 9.98% (weak statistical significance, close to 0,05), while the one of the P100 wave increases almost double, up to 17.88% -highly significant (p<0.001).
Finally, it is observed that all parameters expressing energy, the action power of a wave (the amplitude, duration, area, steepness and IBA) decrease a lot for N75 and increase (although inhibition increases) for P100, even up to 12 or 17.52% ( fig. 5).
In order to have a precise synthetic, numerical information of the balance between excitation and inhibition we realized the ratio between P100/N75 amplitude in hyperthyroid (2.1435) and hypothyroids (2.8713) patients. Therefore, the balance leaned towards inhibition with 33.94% (in favor of inhibition); in hypothyroid patients the same ratios, but for P100/N75 areas, are much higher (3.008 as opposed to 6.098), thus a predomination of inhibition with 108.71% in hypothyroid patients, double compared to hyperthyroid patients (6.098, and 3.008 respectively) (table 12).
Correlations between the parameters of VEP waves and laboratory determinations *T3, T4, RIC, Lipemia etc.). Weak correlations were obtained only between the duration of the N75 wave and all the other 7 non-electrophysiological obtained parameters (Table 13).
The wave duration, on which the surface (energy, power) of the N75 wave depends, shows that it modifies unitarily with the 7 parameters and that these hormonal, humoral, preclinical modifications have an influence on some VEP parameters (duration especially) [22]. It is noticeable that, in all researched literature regarding VEP modifications induced by thyroid pathology, no safe statistical researches are performed through several formulas, nothing is said about the severity of the pathology mirrored in several hormonal, biochemical and paraclinical analyses, and VEP [4]. No correlation calculus is performed between all non-electrophysiological parameters and some electrophysiological parameters -statistical processing that brought a safe base to the existence of a liason of the severity of thyroid pathology with VEP modifications [26].
Plus, all global clinical and VEP researches, in all literature, with no exception, have only followed or shown modifications of the latency of the P100 wave, sometimes of the amplitude and very rarely of the duration [17]. Nothing about the other important waves of VEP (N75 and N135), and measure of all their parameters (latency, amplitude, duration, surface, steepness of wave increase). New, original indices proposed by us, IBA = steepness x amplitude index and LAI = latency/amplitude index, weren't even experimentally determined in literature [25]. Whereas its impossible to make a statement about the balance between excitation (represented by N75) and inhibition (represented by P100), about the mathematic ratio between them and its modification according to thyroid pathology [19].

Conclusions
Following the recording of visual evoked potentials and of 11 humoral and paraclinical parameters on two thyroid patients groups (18 hyperthyroid, 18 hypothyroid), the result were: The age of the illness and treatment duration are much higher in hypothyroid patients (highly significant differences, p<0.01).
The 11 humoral and paraclinical determined parameters, in their majority, have very different values between the 2 groups, with statistically significant differences (p<0.001).
Seven humoral and paraclinical parameters correlate extremely well with each other, (Pearson's r correlation coefficients over 0.9) showing that they modify in parallel with the level of thyroid activity.
The latency of the N75 wave is significantly increased, and the one of the P100 wave is significantly high in hypothyroid patients, with 9.98%, and 17.88% respectively.
The duration of the N75 wave is significantly lower, with statistical significance, in hypothyroid patients, by 26.37%. Short duration = little energy, reduced excitation.
The duration of the N75 wave correlates with most humoral and paraclinical parameters.
The ratios of the amplitudes and of the areas of the P100/ N75 wave indicate a precisely quantified higher predominance of the inhibition in the visual cortex in hypothyroid patients.