Clin Oral Invest DOI 10.1007/s00784-015-1465-9
ORIGINAL ARTICLE
No deficit in neurocortical information processing in patients with cleft lip and palate Volker Gassling 1 & Kai Receveur 1 & Uwe Niederberger 2 & Bernd Koos 3 & Claudia Helene Overath 2 & Falk Birkenfeld 1 & Jörg Wiltfang 1 & Wolf-Dieter Gerber 2 & Michael Siniatchkin 2
Received: 19 October 2014 / Accepted: 18 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Background Cleft lip and palate (CLP) represents the most common malformation of the midfacial region worldwide. It can be suggested that the facial stigmatization, the speech impediment, and the long-standing pressure of treatment cause a range of life stressors. Neurocortical information is influenced by physiological and psychological factors and varies significantly in patients suffering from chronic stress, anxiety, depression, or other psychopathological conditions following maladaptation. The aims of the present study were to investigate the neurocortical information processing of patients with CLP using the contingent negative variation (CNV) paradigm and to evaluate secondary psychopathology, anxiety, and depression. Materials and methods Twenty-five adults with CLP and 25 healthy volunteers, matched in age and gender, were investigated using recordings of the CNV. Initial CNV (iCNV), late CNV (lCNV), and total CNV (tCNV) as well as habituation slope of the iCNV were determined in each subject. Additionally, each participant had to complete the hospital anxiety and
depression questionnaire (HADS) and the Symptom Checklist-90-R (SCL-90-R). Results Individuals with CLP did not differ significantly from healthy subjects according to any of the CNV parameters investigated. No correlations could be revealed between the measured items and the confounding factors age and gender. Additionally, there were no differences between the groups regarding depression and SCL-90-R; however, anxiety showed significant group differences on a subclinical level. Conclusions These data suggest that subjects with CLP show normal neurocortical information processing. It seems likely that CLP and its treatment have no impact on psychosocial functioning and neurophysiological mechanisms of attention. Clinical relevance The specific living conditions of patients with CLP do not result in disease-specific neurophysiological changes. Keywords Event-related potentials . Contingent negative variation . Cleft lip . Psychopathology . Anxiety . Depression
Introduction Trial Registration: German Clinical Trials Register (registration code DRKS00004606). * Volker Gassling [email protected] 1
Department of Oral and Maxillofacial Surgery, University Hospital Schleswig-Holstein, Campus Kiel, Arnold-Heller-Straße 3, Haus 26, 24105 Kiel, Germany
2
Institute of Medical Psychology and Medical Sociology, University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany
3
Department of Orthodontics, University Hospital of SchleswigHolstein, Campus Kiel, Kiel, Germany
Cleft lip and palate (CLP) represents the most common malformation of the midfacial region worldwide [1]. It can be expected that the facial stigmatization, the speech impediment, and the long-standing pressure of treatment cause a range of life stressors which may lead to a long-lasting impact on the affected individuals throughout their lives. A systematic literature review concerning psychosocial stress and CLP reveals that psychosocial functioning among children and young adults with orofacial clefts is affected by behavioral problems, depression, and unhappiness with their facial appearance and speech [2]. In developed countries, psychosocial stress is associated with an increased risk of physical and
Clin Oral Invest
mental health hazards [3]. Disadvantaged population groups are at particular risk. Factors which enhance this risk are social conditions, recurring pressures, and multiple burdens, which lead to chronic stress arousal [4]. Recently, it has been shown that children and adolescents with chronic physical illness show higher levels of depressive symptoms than their unaffected peers, whereby, in addition to chronic fatigue syndrome (d=0.94) and fibromyalgia (d=0.59), the differences were the third strongest in subjects with CLP (d=0.54) [5]. Amazingly, it has also been shown that people with CLP have a higher morbidity throughout their lives (from birth up to the age of 55 years) than unaffected individuals. Particularly evident is a significantly increased risk of cancer, cardiovascular diseases, and suicide [6]. Neurocortical information is influenced by physiological (source localization, neurocortical excitability, neuronal and systemic activation) and psychological factors (motivation, attention, and resource mobilization) and change significantly in patients suffering from chronic stress, anxiety, depression, or other psychopathological conditions following maladaptation [7, 8]. It can be assumed that the particular psychosocial stress influences the neurocortical information processing in patients with CLP. However, until now, very little is known about neurocortical information processing, i.e., neurocortical attention and habituation parameters in CLP-affected individuals. The most common way to investigate neurocortical information processing is to record evoked and event-related potentials (ERPs). The contingent negative variation (CNV) is a slow neurocortical ERP, which can be recorded from the scalp between two stimuli (warning, S1, and imperative stimulus, S2) while the subject is anticipating the second event and preparing for task performance (for example, pressing of a button), i.e., dependent on contingency [7]. From a physiological point of view, the CNV amplitude reflects the excitability of the dendritic trees of neurocortical pyramidal neurons following activation in the cortico-striato-thalamo-cortical loop [9]. If longer (>3 s) interstimulus intervals are employed, subcomponents of the CNV can be separated. The initial CNV component (iCNV, 0.5–1.0 s after S1) has been related to the orienting response [10, 11]. The late CNV component (lCNV, some hundred ms before S2) is believed to represent the stimulus anticipation and response preparation [12, 13]. Both iCNV and lCNV, as well as activity between them, comprise the total CNV (tCNV)—i.e., amplitude between the warning and imperative stimuli. The tCNV is considered to represent overall processes of resource mobilization and executive control over preparation for neuropsychological functions [14]. Thus, CNV is an appropriate method to investigate complex mechanisms of information processing. Moreover, CNV is very sensitive to stress and shows abnormalities (especially decreased amplitudes) in different psychopathologies such as anxiety and depression (see for review Birbaumer et al. [7]). If patients with CLP
are affected by psychopathological disorders, they should demonstrate abnormal CNV amplitudes. Based on the above-mentioned findings, we hypothesized that individuals with CLP firstly would show increased amplitudes and reduced habituation of the iCNV (i.e., dishabituation), and secondly would be characterized by a higher level of psychopathology, anxiety, and depression.
Materials and methods Design A cross-sectional design, comparing a CLP group with ageand gender-matched controls, was employed. Study participants Twenty-five adult patients with a repaired complete unilateral or bilateral non-syndromic CLP (median age 24 years, interquartile range (IQR) 8.5) were recruited from craniofacial malformations consultation appointments with the Department of Cranio-Maxillofacial Surgery, University Hospital of Schleswig-Holstein, Kiel Campus, Kiel, Germany. The control group (matched in age and gender) consists of 25 unaffected volunteers (median age 24 years, IQR 7.87) from Schleswig-Holstein area, who were recruited by newspaper advertisements (Table 1). The minimum and maximum ages were 18 and 51 years in the CLP group and 17 and 52 years in the control group, respectively. The CLP group contained 11 (44 %) females and 14 (56 %) males equivalent to 11 females and 14 males in the control group. They were informed about the course of the experimental procedure and gave their written consent according to the Helsinki convention before examination began. The study design was approved by the local ethics board (AZ: D 456/12) and was registered in the German Clinical Trials Register (registration code: DRKS00004606; https://drks-neu.uniklinik-freiburg.de/drks_web/setLocale_ EN.do). None of the subjects of either group reported mental or psychiatric problems by psychiatric interviews or suffer by tinnitus, hearing disabilities, migraine, Parkinson’s disease, or any other neurological disease. In addition, the intake of any medication is permitted. Table 1 shows the characteristics of the samples. Contingent negative variation recordings The study was performed under laboratory conditions. All participants were lying with eyes open on an armchair located in an electrically shielded room. The subjects were asked not to close, move, or blink their eyes. The CNV was measured by a two-stimulus paradigm [15] using E-Prime 2.0 software for stimulus presentation (Psychology Software Tools, Inc.,
Clin Oral Invest Table 1 Demographic characteristics of the CLP group and the control group, matched in age and gender. The nonparametric Mann-WhitneyWilcoxon test for comparison of two independent groups shows the level of significance
Age (years) Median IQR Range Gender w m
CLP group N=25
Control group N=25
Test of significance Mann-Whitney-Wilcoxon test
24 8.5 18–51
24 7.87 17–52
p=0.5789
11 (44 %) 14 (56 %)
11 (44 %) 14 (56 %)
IQR interquartile range
Sharpsburg, USA). The auditory warning (S1, f=1000 Hz, 75 dB(a), 100 ms) and imperative (S2, f = 2500 Hz, 75 dB(a), maximal 3000 ms) stimuli were produced by a loudspeaker located before the subject, with an interstimulus interval of 3 s. A CNV session consisted of 32 trials in which the subject had to react to the imperative stimulus (go response) by pressing a button. Additionally, eight trials were randomly presented where no reaction was expected (no-go response—S3, f=200 Hz, 75 dB(a), 100 ms). The time constant between two trials varied between 6 and 10 s. The duration of one recording was 5 s (the recording began 1 s before S1 and ended 1 s after S2). The period between recording onset and S1 was taken as the baseline for all measurements. The EEG was recorded with a 32-channel Easycap (Ag/ AgCl ring electrodes, Falk Minow Co.) using BrainVision MR amplifiers (Brain Products Co., Gilching, Germany) and Brain Vision Recorder (version 1.05) software. The recordings were conducted in accordance with the international 1020 system described by Jasper [16], whereby EEG was recorded over Fz, Cz, and Pz. The electrodes were connected with the EEG shower and impedances were tested. Reference was located at the left mastoid and grounding was located behind the right ear. The electrode impedance was below 10 kΩ. The sampling rate was 250 Hz/s, the high-pass filter was 250 Hz, and the low-pass filter had a time constant of 10 s. Vertical eye movement artifacts were excluded by parallel recording of the vertical electrooculogram (VEOG). The VEOG was located below the right eye. The trial was rejected if EOG deflections were greater than 20 mV. There were no significant differences between the groups according to the number of rejected trials. CNV analyses Raw data were analyzed by Brain Vision Analyzer 1 (Brain Products Co.). EEG signals were filtered offline by a highpass filter (30 Hz, 13 dB/oct) and a notch filter (50 Hz). After segmentation, baseline correction was made for intervals 1000 ms before S1. Ocular correction was undertaken in accordance with the method of Gratton and Coles [17]. Trials
with movement or EOG artifacts, not registered by the VEOG (−100 and +100 μV from 500 ms before and after a blink), were rejected. The 32 go trails were averaged. The total CNV (tCNV) was assessed between 0 and 3000 ms following S1. The initial CNV (iCNV) component was defined as the mean amplitude in a window of 200 ms duration around the maximal amplitude of the expectancy wave between 550 and 750 ms after S1 [18]. The late component of the CNV (lCNV) was assessed as the mean amplitude during the 200 ms preceding S2. To determine the course of habituation, each recording was divided into eight blocks. Each block was built out of four go trials. Habituation was indicated by a positive slope (hence a decrease of the regression line), whereas dishabituation was marked by a negative slope (hence an increase of the regression line) calculated by linear regression (y=ax+b, where a is the slope of habituation) [19]. This study considered only the early CNV component habituation data. This limitation of data presentation is based on the literature [19–25], whereby only the early CNV showed habituation differences between the groups of migraine patients and healthy controls.
Psychological assessment Two different questionnaires were applied, firstly, the German version of the English Hospital Anxiety and Depression Scale (HADS) which was denoted as Hospital Anxiety and Depression Scale-D (HADS-D). This questionnaire is a screening method for the assessment of anxiety and depression. The HADS was designed by Zigmond and Snaith in 1983 [26]. The German form, HADS-D, which was developed by Herrmann et al. [27], contains 14 statements which were summarized to the subscales anxiety and depression with 7 items each. The existence or expression of symptoms during the last week has to be assessed on a four-stage rating scale. The achieved values were brought into relation with cutoff values. The internal consistencies (Cronbach’s α) were α=0.80 for the HADS-D/A and α=0.81 for the HADS-D/D. This demonstrates a good statistical reliability of the scales.
Clin Oral Invest
Secondly, the Symptom Checklist-90-R (SCL-90-R) serves as self-assessment procedure to detect various burdening symptoms. The SCL-90-R closes the gap between temporal extremely variable sensitivity and temporal outlasting personality structure, by measuring the subjective impairment with 90 predefined physical and mental symptoms within the timeframe of 7 days. The SCL-90-R was designed by Derogatis in 1977 and translated to German in 1981 [28]. The test contains 90 items which represent different symptoms. According to the 90 statements, answers were divided into 9 subscales, nominally somatization, obsessivecompulsive disorder, interpersonal, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. Each item has to be answered on a five-step Likert scale. The level of disability was measured as follows: Bnever^=0; Ba bit^=1; Bquite^=2; Bstrong^=3; and Bvery strong^=4. The internal consistencies (Cronbach’s α) for the collective proceedings was α=0.97. This demonstrates a good statistical reliability of the scales.
Data transformation and statistical analysis Statistical analyses were executed by SPSS 20 with a significant level of 5 % for all tests. The non-parametric MannWhitney-Wilcoxon test for comparison of two independent groups was chosen for the calculation of differences in the demographic characteristics of the study group and to test the influence of the confounding factor gender on the measured items. The amount of correlation of the age, in terms of
Fig. 1 The figure shows the amplitudes of the initial, late, and total CNV and reaction time (RT) for the CLP group and the control group. (RT=reaction time/SD= standard deviation)
confounding influence, was analyzed by Spearman’s rank correlation coefficient. The CNV and psychological assessment item scores were evaluated using the t test for independent samples. All data were tested before for normal distribution via Kolmogorov-Smirnov test and for homogeneity of variance via Levene test.
Results CNV amplitudes Figure 1 shows the amplitudes of the initial, late, and total CNV and reaction times, and Table 2 demonstrates the corresponding values (means and standard deviations) for subjects with CLP and control individuals. Both groups showed similar CNV amplitudes. On an average, the healthy control group demonstrated smaller CNV amplitudes. Despite the means of the control group having been less negative compared to those of the CLP group, the statistical significance level has not been achieved. Concerning the reaction time, there was no significant difference found between the groups. CNV habituation Figure 2 shows habituation characteristics of the CLP group and the control group. Both the CLP group and healthy control group showed a negative slope (Table 3). As the negative slope reflects dishabituation, both groups showed CNV
Clin Oral Invest Table 2 Mean and standard deviations of total CNV (tCNV), initial CNV (iCNV), and late CNV (lCNV), with reference to the CLP group or the control group. The t test shows the level of significance CLP group
Total CNV Mean SD Initial CNV Mean SD Late CNV Mean SD Reaction time Mean SD
Control group
Test of significance t test
Table 3 Comparison of mean slope a according to the y=ax+b as habituation parameter for the CLP group and control group Slope of habituation CLP group Control group Test of significance t test −0.32 0.86
Mean SD −5.71 3.36
−4.31 2.05
t(48)=−1.786 p=0.080
−6.68 4.43
−5.95 3.9
t(48)=−0.619 p=0.539
−8.78 5.55
−7.19 4.32
t(48)=−1.127 p=0.265
299.79 57.80
292.50 53.40
t(48)=0.463 p=0.645
−0.22 1.42
t(48)=−0.303 p=0.763
Confounding factors Concerning possible influences of gender and age on the above-mentioned measured items, the Mann-WhitneyWilcoxon test showed no statistically significant differences between gender and the various items (p between 0.1708 and 0.9899). Spearmen’s correlation coefficient, evaluating the influence of age, revealed a mean of 0.04 (std. dev. 0.22, range between −0.36 and 0.41), not showing concrete correlations between age and the measured items.
characteristics. The groups did not differ significantly according to habituation (t(48)=−0.303).
Discussion Evaluation of questionnaires Averaging, the items of the control group were smaller compared to those of the CLP group. Concerning the value of anxiety, the control group was 3.04 and the CLP group 5.04. On a subclinical level (scores < 10), the t test shows a significant difference (p = 0.027). The value of depression showed no significance (p = 0.272). Altogether, both groups showed negative items (0–7) according to the definition of the HADS questionnaire. This is why an outstanding presence of anxiety and depression can be excluded. In addition, the SCL90-R was also analyzed with regard to the mean, standard deviation, and clinical significance. On the whole, there were no significant differences between the groups according to parameters of the SCL-90-R (Table 4). -12 CLP Controls
Amplitude Cz (uV)
Fig. 2 Habituation of the initial CNV component represented as a course of amplitudes over eight blocks of recording or a slope calculated by linear regression (y=ax+b) for subjects with CLP and control subjects. Coefficients of linear regression (a) are shown for each group in the legend box. (cz=electrode; CLP group, a= −0.32; control group, a=−0.22)
The aim of the present exploratory study was to investigate the neurocortical information processing and the level of psychopathology, anxiety, and depression in individuals with CLP. We hypothesized that CLP-affected individuals firstly would show increased amplitudes and reduced habituation of the iCNV (i.e., dishabituation), and secondly would be characterized by a higher level of psychopathology, anxiety, and depression. Against our hypothesis, the main findings of our study are as follows: 1) the evaluation of the CNV revealed no significant differences between CLP-affected individuals and control subjects for amplitudes and habituation of the iCNV and 2) the analysis of psychopathology, anxiety, and depression revealed no significant group differences for SCL-90-R and depression, whereby anxiety showed significant group differences on a subclinical level.
-10
-8
-6
Block 1 Block 2 Block 3 Block 4 Block 5 Block 6 Block 7 Block 8
Clin Oral Invest Table 4 Mean and standard deviations of HADS-D and SCL-90-R questionnaires with reference to the CLP group or the control group. The t test shows the level of significance CLP group
Control group
HADS-D anxiety Mean 5.04 3.04 SD 2.94 3.25 HADS-D depression Mean 2.36 1.64 SD 2.34 2.23 SCL-90-R somatization Mean 48.00 49.40 SD 8.66 9.20 SCL-90-R obsessive-compulsive Mean 50.28 48.44 SD 10.78 9.50 SCL-90-R interpersonal Mean 48.16 44.52 SD 11.14 SCL-90-R depression Mean 51.20 SD 10.87 SCL-90-R anxiety Mean 46.40 SD 8.35 SCL-90-R hostility Mean 46.00 SD 0.0 SCL-90-R phobic anxiety Mean 47.52 SD 9.17 SCL-90-R paranoid ideation Mean 49.16 SD 11.70 SCL-90-R psychoticism Mean 49.24 SD
10.15
Test of significance t test
t(48)=2.284 p=0.027 t(48)=1.112 p=0.272 t(48)=−0.554 p=0.582 t(48)=0.640 p=0.525 t(48)=1.374
7.18
p=0.176
47.84 9.38
t(48)=1.170 p=0.248
46.24 5.83
t(48)=0.079 p=0.938
46.00 0.0
* *
44.80 6.81
t(48)=1.191 p=0.239
46.20 6.69
t(48)=1.098 p=0.278
46.72
t(48)=1.080
5.76
p=0.286
*Due to identical distribution of measured values no p-value and t(48) can be calculated within the t-test
As mentioned above, it is known that ERPs are influenced by psychological and physiological factors [7, 8]. Individuals born with a facial disfigurement are forced to undergo multiple surgical procedures and frequent clinic attendances from birth to adulthood. The result is a more or less intensive disturbance of facial appearance and speech impediment which might lead to particular psychosocial stress in affected individuals. More precisely, dissatisfaction with facial appearance could be localized to the upper lip, and the nose and nasal breathing. Furthermore, even a positive correlation between
satisfaction with facial appearance and a health-related quality of life has been found in adults with bilateral CLP [29]. Several years ago, it was shown that the severity of the cleft deformity may have a significant impact on social competence in childhood, e.g., in the development of friendships. From this, it is not surprising that even adult siblings with repaired clefts were shown to be less likely to marry than their non-cleft siblings [30]. Recently, these findings have been explained by a comparison of a visual assessment of faces and it was found that in CLP faces, there were more initial fixations in the mouth and longer fixations in the mouth and nose regions compared to control faces. In addition, CLP faces were rated more negatively overall [31]. Even speech problems have been found to affect the psychosocial adjustment of individuals with clefts of the orofacial region. Recently, these findings have been confirmed in adolescents who have reported speech problems as an important factor in psychosocial adjustment difficulties [32]. Amazingly, it has been found that speech problems, and not more or less abnormal facial appearance, lead to better psychosocial health in adolescents [33]. On the other hand, it is known that an important indicator of psychosocial adjustment difficulties is the occurrence of teasing in childhood. In this regard, several studies have pointed out a positive correlation between the conspicuous lip appearance and speech problems [34, 35]. To summarize the abovementioned findings, it can be assumed that concerns about facial appearance and speech are the major features of psychosocial adjustment difficulties in CLP-related malformations. It can be assumed that this particular psychosocial stress may alter the neurophysiological development of affected individuals. However, against our first hypothesis, the results of the present study revealed that subjects with CLP showed no neurophysiological alterations, i.e., increased amplitudes and reduced habituation (i.e., dishabituation) of the iCNV. These findings are in line with recent findings of our group concerning stress coping in CLP patients. Here, it could be shown that adults with CLP show significant better endocrinological stress coping than healthy peers [36]. It is conceivable that the lives of CLP-affected subjects, with their special challenges and burdens, result in resilience [37]. In general, it is supposed that individuals with major life stressors in their history can benefit in the sense of positive outcomes [38]. In this context, it has been shown that children with CLP showed good resilience with adequate emotional functioning, high satisfaction with appearance, and lower frequency of reported teasing. In contrast, it has been shown that the child’s characteristics such as visibility of cleft, gender, and additional diagnosis do not generally contribute to the explanation of psychosocial resilience [39]. The above-mentioned aspects may also be an explanation for the normal neurocortical attention and habituation parameters of individuals with CLP. Concerning the second hypothesis, there is increased anxiety of CLP patients in comparison to the control group;
Clin Oral Invest
however, differences are on a subclinical level (scores
ORIGINAL ARTICLE
No deficit in neurocortical information processing in patients with cleft lip and palate Volker Gassling 1 & Kai Receveur 1 & Uwe Niederberger 2 & Bernd Koos 3 & Claudia Helene Overath 2 & Falk Birkenfeld 1 & Jörg Wiltfang 1 & Wolf-Dieter Gerber 2 & Michael Siniatchkin 2
Received: 19 October 2014 / Accepted: 18 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Background Cleft lip and palate (CLP) represents the most common malformation of the midfacial region worldwide. It can be suggested that the facial stigmatization, the speech impediment, and the long-standing pressure of treatment cause a range of life stressors. Neurocortical information is influenced by physiological and psychological factors and varies significantly in patients suffering from chronic stress, anxiety, depression, or other psychopathological conditions following maladaptation. The aims of the present study were to investigate the neurocortical information processing of patients with CLP using the contingent negative variation (CNV) paradigm and to evaluate secondary psychopathology, anxiety, and depression. Materials and methods Twenty-five adults with CLP and 25 healthy volunteers, matched in age and gender, were investigated using recordings of the CNV. Initial CNV (iCNV), late CNV (lCNV), and total CNV (tCNV) as well as habituation slope of the iCNV were determined in each subject. Additionally, each participant had to complete the hospital anxiety and
depression questionnaire (HADS) and the Symptom Checklist-90-R (SCL-90-R). Results Individuals with CLP did not differ significantly from healthy subjects according to any of the CNV parameters investigated. No correlations could be revealed between the measured items and the confounding factors age and gender. Additionally, there were no differences between the groups regarding depression and SCL-90-R; however, anxiety showed significant group differences on a subclinical level. Conclusions These data suggest that subjects with CLP show normal neurocortical information processing. It seems likely that CLP and its treatment have no impact on psychosocial functioning and neurophysiological mechanisms of attention. Clinical relevance The specific living conditions of patients with CLP do not result in disease-specific neurophysiological changes. Keywords Event-related potentials . Contingent negative variation . Cleft lip . Psychopathology . Anxiety . Depression
Introduction Trial Registration: German Clinical Trials Register (registration code DRKS00004606). * Volker Gassling [email protected] 1
Department of Oral and Maxillofacial Surgery, University Hospital Schleswig-Holstein, Campus Kiel, Arnold-Heller-Straße 3, Haus 26, 24105 Kiel, Germany
2
Institute of Medical Psychology and Medical Sociology, University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany
3
Department of Orthodontics, University Hospital of SchleswigHolstein, Campus Kiel, Kiel, Germany
Cleft lip and palate (CLP) represents the most common malformation of the midfacial region worldwide [1]. It can be expected that the facial stigmatization, the speech impediment, and the long-standing pressure of treatment cause a range of life stressors which may lead to a long-lasting impact on the affected individuals throughout their lives. A systematic literature review concerning psychosocial stress and CLP reveals that psychosocial functioning among children and young adults with orofacial clefts is affected by behavioral problems, depression, and unhappiness with their facial appearance and speech [2]. In developed countries, psychosocial stress is associated with an increased risk of physical and
Clin Oral Invest
mental health hazards [3]. Disadvantaged population groups are at particular risk. Factors which enhance this risk are social conditions, recurring pressures, and multiple burdens, which lead to chronic stress arousal [4]. Recently, it has been shown that children and adolescents with chronic physical illness show higher levels of depressive symptoms than their unaffected peers, whereby, in addition to chronic fatigue syndrome (d=0.94) and fibromyalgia (d=0.59), the differences were the third strongest in subjects with CLP (d=0.54) [5]. Amazingly, it has also been shown that people with CLP have a higher morbidity throughout their lives (from birth up to the age of 55 years) than unaffected individuals. Particularly evident is a significantly increased risk of cancer, cardiovascular diseases, and suicide [6]. Neurocortical information is influenced by physiological (source localization, neurocortical excitability, neuronal and systemic activation) and psychological factors (motivation, attention, and resource mobilization) and change significantly in patients suffering from chronic stress, anxiety, depression, or other psychopathological conditions following maladaptation [7, 8]. It can be assumed that the particular psychosocial stress influences the neurocortical information processing in patients with CLP. However, until now, very little is known about neurocortical information processing, i.e., neurocortical attention and habituation parameters in CLP-affected individuals. The most common way to investigate neurocortical information processing is to record evoked and event-related potentials (ERPs). The contingent negative variation (CNV) is a slow neurocortical ERP, which can be recorded from the scalp between two stimuli (warning, S1, and imperative stimulus, S2) while the subject is anticipating the second event and preparing for task performance (for example, pressing of a button), i.e., dependent on contingency [7]. From a physiological point of view, the CNV amplitude reflects the excitability of the dendritic trees of neurocortical pyramidal neurons following activation in the cortico-striato-thalamo-cortical loop [9]. If longer (>3 s) interstimulus intervals are employed, subcomponents of the CNV can be separated. The initial CNV component (iCNV, 0.5–1.0 s after S1) has been related to the orienting response [10, 11]. The late CNV component (lCNV, some hundred ms before S2) is believed to represent the stimulus anticipation and response preparation [12, 13]. Both iCNV and lCNV, as well as activity between them, comprise the total CNV (tCNV)—i.e., amplitude between the warning and imperative stimuli. The tCNV is considered to represent overall processes of resource mobilization and executive control over preparation for neuropsychological functions [14]. Thus, CNV is an appropriate method to investigate complex mechanisms of information processing. Moreover, CNV is very sensitive to stress and shows abnormalities (especially decreased amplitudes) in different psychopathologies such as anxiety and depression (see for review Birbaumer et al. [7]). If patients with CLP
are affected by psychopathological disorders, they should demonstrate abnormal CNV amplitudes. Based on the above-mentioned findings, we hypothesized that individuals with CLP firstly would show increased amplitudes and reduced habituation of the iCNV (i.e., dishabituation), and secondly would be characterized by a higher level of psychopathology, anxiety, and depression.
Materials and methods Design A cross-sectional design, comparing a CLP group with ageand gender-matched controls, was employed. Study participants Twenty-five adult patients with a repaired complete unilateral or bilateral non-syndromic CLP (median age 24 years, interquartile range (IQR) 8.5) were recruited from craniofacial malformations consultation appointments with the Department of Cranio-Maxillofacial Surgery, University Hospital of Schleswig-Holstein, Kiel Campus, Kiel, Germany. The control group (matched in age and gender) consists of 25 unaffected volunteers (median age 24 years, IQR 7.87) from Schleswig-Holstein area, who were recruited by newspaper advertisements (Table 1). The minimum and maximum ages were 18 and 51 years in the CLP group and 17 and 52 years in the control group, respectively. The CLP group contained 11 (44 %) females and 14 (56 %) males equivalent to 11 females and 14 males in the control group. They were informed about the course of the experimental procedure and gave their written consent according to the Helsinki convention before examination began. The study design was approved by the local ethics board (AZ: D 456/12) and was registered in the German Clinical Trials Register (registration code: DRKS00004606; https://drks-neu.uniklinik-freiburg.de/drks_web/setLocale_ EN.do). None of the subjects of either group reported mental or psychiatric problems by psychiatric interviews or suffer by tinnitus, hearing disabilities, migraine, Parkinson’s disease, or any other neurological disease. In addition, the intake of any medication is permitted. Table 1 shows the characteristics of the samples. Contingent negative variation recordings The study was performed under laboratory conditions. All participants were lying with eyes open on an armchair located in an electrically shielded room. The subjects were asked not to close, move, or blink their eyes. The CNV was measured by a two-stimulus paradigm [15] using E-Prime 2.0 software for stimulus presentation (Psychology Software Tools, Inc.,
Clin Oral Invest Table 1 Demographic characteristics of the CLP group and the control group, matched in age and gender. The nonparametric Mann-WhitneyWilcoxon test for comparison of two independent groups shows the level of significance
Age (years) Median IQR Range Gender w m
CLP group N=25
Control group N=25
Test of significance Mann-Whitney-Wilcoxon test
24 8.5 18–51
24 7.87 17–52
p=0.5789
11 (44 %) 14 (56 %)
11 (44 %) 14 (56 %)
IQR interquartile range
Sharpsburg, USA). The auditory warning (S1, f=1000 Hz, 75 dB(a), 100 ms) and imperative (S2, f = 2500 Hz, 75 dB(a), maximal 3000 ms) stimuli were produced by a loudspeaker located before the subject, with an interstimulus interval of 3 s. A CNV session consisted of 32 trials in which the subject had to react to the imperative stimulus (go response) by pressing a button. Additionally, eight trials were randomly presented where no reaction was expected (no-go response—S3, f=200 Hz, 75 dB(a), 100 ms). The time constant between two trials varied between 6 and 10 s. The duration of one recording was 5 s (the recording began 1 s before S1 and ended 1 s after S2). The period between recording onset and S1 was taken as the baseline for all measurements. The EEG was recorded with a 32-channel Easycap (Ag/ AgCl ring electrodes, Falk Minow Co.) using BrainVision MR amplifiers (Brain Products Co., Gilching, Germany) and Brain Vision Recorder (version 1.05) software. The recordings were conducted in accordance with the international 1020 system described by Jasper [16], whereby EEG was recorded over Fz, Cz, and Pz. The electrodes were connected with the EEG shower and impedances were tested. Reference was located at the left mastoid and grounding was located behind the right ear. The electrode impedance was below 10 kΩ. The sampling rate was 250 Hz/s, the high-pass filter was 250 Hz, and the low-pass filter had a time constant of 10 s. Vertical eye movement artifacts were excluded by parallel recording of the vertical electrooculogram (VEOG). The VEOG was located below the right eye. The trial was rejected if EOG deflections were greater than 20 mV. There were no significant differences between the groups according to the number of rejected trials. CNV analyses Raw data were analyzed by Brain Vision Analyzer 1 (Brain Products Co.). EEG signals were filtered offline by a highpass filter (30 Hz, 13 dB/oct) and a notch filter (50 Hz). After segmentation, baseline correction was made for intervals 1000 ms before S1. Ocular correction was undertaken in accordance with the method of Gratton and Coles [17]. Trials
with movement or EOG artifacts, not registered by the VEOG (−100 and +100 μV from 500 ms before and after a blink), were rejected. The 32 go trails were averaged. The total CNV (tCNV) was assessed between 0 and 3000 ms following S1. The initial CNV (iCNV) component was defined as the mean amplitude in a window of 200 ms duration around the maximal amplitude of the expectancy wave between 550 and 750 ms after S1 [18]. The late component of the CNV (lCNV) was assessed as the mean amplitude during the 200 ms preceding S2. To determine the course of habituation, each recording was divided into eight blocks. Each block was built out of four go trials. Habituation was indicated by a positive slope (hence a decrease of the regression line), whereas dishabituation was marked by a negative slope (hence an increase of the regression line) calculated by linear regression (y=ax+b, where a is the slope of habituation) [19]. This study considered only the early CNV component habituation data. This limitation of data presentation is based on the literature [19–25], whereby only the early CNV showed habituation differences between the groups of migraine patients and healthy controls.
Psychological assessment Two different questionnaires were applied, firstly, the German version of the English Hospital Anxiety and Depression Scale (HADS) which was denoted as Hospital Anxiety and Depression Scale-D (HADS-D). This questionnaire is a screening method for the assessment of anxiety and depression. The HADS was designed by Zigmond and Snaith in 1983 [26]. The German form, HADS-D, which was developed by Herrmann et al. [27], contains 14 statements which were summarized to the subscales anxiety and depression with 7 items each. The existence or expression of symptoms during the last week has to be assessed on a four-stage rating scale. The achieved values were brought into relation with cutoff values. The internal consistencies (Cronbach’s α) were α=0.80 for the HADS-D/A and α=0.81 for the HADS-D/D. This demonstrates a good statistical reliability of the scales.
Clin Oral Invest
Secondly, the Symptom Checklist-90-R (SCL-90-R) serves as self-assessment procedure to detect various burdening symptoms. The SCL-90-R closes the gap between temporal extremely variable sensitivity and temporal outlasting personality structure, by measuring the subjective impairment with 90 predefined physical and mental symptoms within the timeframe of 7 days. The SCL-90-R was designed by Derogatis in 1977 and translated to German in 1981 [28]. The test contains 90 items which represent different symptoms. According to the 90 statements, answers were divided into 9 subscales, nominally somatization, obsessivecompulsive disorder, interpersonal, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. Each item has to be answered on a five-step Likert scale. The level of disability was measured as follows: Bnever^=0; Ba bit^=1; Bquite^=2; Bstrong^=3; and Bvery strong^=4. The internal consistencies (Cronbach’s α) for the collective proceedings was α=0.97. This demonstrates a good statistical reliability of the scales.
Data transformation and statistical analysis Statistical analyses were executed by SPSS 20 with a significant level of 5 % for all tests. The non-parametric MannWhitney-Wilcoxon test for comparison of two independent groups was chosen for the calculation of differences in the demographic characteristics of the study group and to test the influence of the confounding factor gender on the measured items. The amount of correlation of the age, in terms of
Fig. 1 The figure shows the amplitudes of the initial, late, and total CNV and reaction time (RT) for the CLP group and the control group. (RT=reaction time/SD= standard deviation)
confounding influence, was analyzed by Spearman’s rank correlation coefficient. The CNV and psychological assessment item scores were evaluated using the t test for independent samples. All data were tested before for normal distribution via Kolmogorov-Smirnov test and for homogeneity of variance via Levene test.
Results CNV amplitudes Figure 1 shows the amplitudes of the initial, late, and total CNV and reaction times, and Table 2 demonstrates the corresponding values (means and standard deviations) for subjects with CLP and control individuals. Both groups showed similar CNV amplitudes. On an average, the healthy control group demonstrated smaller CNV amplitudes. Despite the means of the control group having been less negative compared to those of the CLP group, the statistical significance level has not been achieved. Concerning the reaction time, there was no significant difference found between the groups. CNV habituation Figure 2 shows habituation characteristics of the CLP group and the control group. Both the CLP group and healthy control group showed a negative slope (Table 3). As the negative slope reflects dishabituation, both groups showed CNV
Clin Oral Invest Table 2 Mean and standard deviations of total CNV (tCNV), initial CNV (iCNV), and late CNV (lCNV), with reference to the CLP group or the control group. The t test shows the level of significance CLP group
Total CNV Mean SD Initial CNV Mean SD Late CNV Mean SD Reaction time Mean SD
Control group
Test of significance t test
Table 3 Comparison of mean slope a according to the y=ax+b as habituation parameter for the CLP group and control group Slope of habituation CLP group Control group Test of significance t test −0.32 0.86
Mean SD −5.71 3.36
−4.31 2.05
t(48)=−1.786 p=0.080
−6.68 4.43
−5.95 3.9
t(48)=−0.619 p=0.539
−8.78 5.55
−7.19 4.32
t(48)=−1.127 p=0.265
299.79 57.80
292.50 53.40
t(48)=0.463 p=0.645
−0.22 1.42
t(48)=−0.303 p=0.763
Confounding factors Concerning possible influences of gender and age on the above-mentioned measured items, the Mann-WhitneyWilcoxon test showed no statistically significant differences between gender and the various items (p between 0.1708 and 0.9899). Spearmen’s correlation coefficient, evaluating the influence of age, revealed a mean of 0.04 (std. dev. 0.22, range between −0.36 and 0.41), not showing concrete correlations between age and the measured items.
characteristics. The groups did not differ significantly according to habituation (t(48)=−0.303).
Discussion Evaluation of questionnaires Averaging, the items of the control group were smaller compared to those of the CLP group. Concerning the value of anxiety, the control group was 3.04 and the CLP group 5.04. On a subclinical level (scores < 10), the t test shows a significant difference (p = 0.027). The value of depression showed no significance (p = 0.272). Altogether, both groups showed negative items (0–7) according to the definition of the HADS questionnaire. This is why an outstanding presence of anxiety and depression can be excluded. In addition, the SCL90-R was also analyzed with regard to the mean, standard deviation, and clinical significance. On the whole, there were no significant differences between the groups according to parameters of the SCL-90-R (Table 4). -12 CLP Controls
Amplitude Cz (uV)
Fig. 2 Habituation of the initial CNV component represented as a course of amplitudes over eight blocks of recording or a slope calculated by linear regression (y=ax+b) for subjects with CLP and control subjects. Coefficients of linear regression (a) are shown for each group in the legend box. (cz=electrode; CLP group, a= −0.32; control group, a=−0.22)
The aim of the present exploratory study was to investigate the neurocortical information processing and the level of psychopathology, anxiety, and depression in individuals with CLP. We hypothesized that CLP-affected individuals firstly would show increased amplitudes and reduced habituation of the iCNV (i.e., dishabituation), and secondly would be characterized by a higher level of psychopathology, anxiety, and depression. Against our hypothesis, the main findings of our study are as follows: 1) the evaluation of the CNV revealed no significant differences between CLP-affected individuals and control subjects for amplitudes and habituation of the iCNV and 2) the analysis of psychopathology, anxiety, and depression revealed no significant group differences for SCL-90-R and depression, whereby anxiety showed significant group differences on a subclinical level.
-10
-8
-6
Block 1 Block 2 Block 3 Block 4 Block 5 Block 6 Block 7 Block 8
Clin Oral Invest Table 4 Mean and standard deviations of HADS-D and SCL-90-R questionnaires with reference to the CLP group or the control group. The t test shows the level of significance CLP group
Control group
HADS-D anxiety Mean 5.04 3.04 SD 2.94 3.25 HADS-D depression Mean 2.36 1.64 SD 2.34 2.23 SCL-90-R somatization Mean 48.00 49.40 SD 8.66 9.20 SCL-90-R obsessive-compulsive Mean 50.28 48.44 SD 10.78 9.50 SCL-90-R interpersonal Mean 48.16 44.52 SD 11.14 SCL-90-R depression Mean 51.20 SD 10.87 SCL-90-R anxiety Mean 46.40 SD 8.35 SCL-90-R hostility Mean 46.00 SD 0.0 SCL-90-R phobic anxiety Mean 47.52 SD 9.17 SCL-90-R paranoid ideation Mean 49.16 SD 11.70 SCL-90-R psychoticism Mean 49.24 SD
10.15
Test of significance t test
t(48)=2.284 p=0.027 t(48)=1.112 p=0.272 t(48)=−0.554 p=0.582 t(48)=0.640 p=0.525 t(48)=1.374
7.18
p=0.176
47.84 9.38
t(48)=1.170 p=0.248
46.24 5.83
t(48)=0.079 p=0.938
46.00 0.0
* *
44.80 6.81
t(48)=1.191 p=0.239
46.20 6.69
t(48)=1.098 p=0.278
46.72
t(48)=1.080
5.76
p=0.286
*Due to identical distribution of measured values no p-value and t(48) can be calculated within the t-test
As mentioned above, it is known that ERPs are influenced by psychological and physiological factors [7, 8]. Individuals born with a facial disfigurement are forced to undergo multiple surgical procedures and frequent clinic attendances from birth to adulthood. The result is a more or less intensive disturbance of facial appearance and speech impediment which might lead to particular psychosocial stress in affected individuals. More precisely, dissatisfaction with facial appearance could be localized to the upper lip, and the nose and nasal breathing. Furthermore, even a positive correlation between
satisfaction with facial appearance and a health-related quality of life has been found in adults with bilateral CLP [29]. Several years ago, it was shown that the severity of the cleft deformity may have a significant impact on social competence in childhood, e.g., in the development of friendships. From this, it is not surprising that even adult siblings with repaired clefts were shown to be less likely to marry than their non-cleft siblings [30]. Recently, these findings have been explained by a comparison of a visual assessment of faces and it was found that in CLP faces, there were more initial fixations in the mouth and longer fixations in the mouth and nose regions compared to control faces. In addition, CLP faces were rated more negatively overall [31]. Even speech problems have been found to affect the psychosocial adjustment of individuals with clefts of the orofacial region. Recently, these findings have been confirmed in adolescents who have reported speech problems as an important factor in psychosocial adjustment difficulties [32]. Amazingly, it has been found that speech problems, and not more or less abnormal facial appearance, lead to better psychosocial health in adolescents [33]. On the other hand, it is known that an important indicator of psychosocial adjustment difficulties is the occurrence of teasing in childhood. In this regard, several studies have pointed out a positive correlation between the conspicuous lip appearance and speech problems [34, 35]. To summarize the abovementioned findings, it can be assumed that concerns about facial appearance and speech are the major features of psychosocial adjustment difficulties in CLP-related malformations. It can be assumed that this particular psychosocial stress may alter the neurophysiological development of affected individuals. However, against our first hypothesis, the results of the present study revealed that subjects with CLP showed no neurophysiological alterations, i.e., increased amplitudes and reduced habituation (i.e., dishabituation) of the iCNV. These findings are in line with recent findings of our group concerning stress coping in CLP patients. Here, it could be shown that adults with CLP show significant better endocrinological stress coping than healthy peers [36]. It is conceivable that the lives of CLP-affected subjects, with their special challenges and burdens, result in resilience [37]. In general, it is supposed that individuals with major life stressors in their history can benefit in the sense of positive outcomes [38]. In this context, it has been shown that children with CLP showed good resilience with adequate emotional functioning, high satisfaction with appearance, and lower frequency of reported teasing. In contrast, it has been shown that the child’s characteristics such as visibility of cleft, gender, and additional diagnosis do not generally contribute to the explanation of psychosocial resilience [39]. The above-mentioned aspects may also be an explanation for the normal neurocortical attention and habituation parameters of individuals with CLP. Concerning the second hypothesis, there is increased anxiety of CLP patients in comparison to the control group;
Clin Oral Invest
however, differences are on a subclinical level (scores
