Graefes Arch Clin Exp Ophthalmol DOI 10.1007/s00417-015-2981-7
NEUROPHTHALMOLOGY
Peripapillary retinal nerve fiber layer thickness in bipolar disorder Ali Mehraban 1 & Seyed Mehdi Samimi 1 & Morteza Entezari 2 & Mohammad Hassan Seifi 2 & Maryam Nazari 3 & Mehdi Yaseri 4
Received: 13 June 2014 / Revised: 12 December 2014 / Accepted: 2 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Background To compare peripapillary retinal nerve fiber layer thickness (RNFLT) between patients with bipolar disorder and a control group by optical coherence tomography (OCT). Methods This prospective comparative case series included 60 eyes of 30 patients with bipolar disorder and 60 eyes of The work was carried out by Ophthalmic Research Centre Shahid Beheshti Medical University. Ali Mehraban is a psychiatrist, Psychiatry Research Center, Department of Psychiatry, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Seyed Mehdi Samimi is an associate professor of Psychiatry, Psychiatry Research Center, Department of Psychiatry, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Morteza Entezari is an associate professor of Ophthalmology, Ophthalmic Research Center, Department of Ophthalmology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Mohammad Hassan Seifi is an ophthalmologist, Ophthalmic Research Center, Department of Ophthalmology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Maryam Nazary is a neurologist, Department of Neurology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Mehdi Yaseri is a biostatistician, Department of Epidemiology and Biostatistics, Tehran University of Medical Sciences, and holds a PhD. * Morteza Entezari [email protected] 1
Psychiatry Research Center, Department of Psychiatry, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran
2
Ophthalmic Research Center, Department of Ophthalmology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, Shahid Madani Ave., 1617763141 Tehran, Iran
3
Department of Neurology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran
4
Department of Epidemiology and Biostatistics, Tehran University of Medical Sciences, Tehran, Iran
30 age-matched healthy control subjects. Using OCT, peripapillary RNFLT of the 4 quadrants and the mean of them was compared between the two groups. Variables such as age of onset, duration, smoking, psychosis, mania and depression episodes in the case group and their relationships with RNFLT were evaluated by OCT. Results Mean RNFLT was 99±8 in the case group, significantly less than the106±8 mμ in the control group (p=0.001). The inferior, superior, and nasal quadrants in the case and control groups showed significant difference in RNFLT (p < 0.001) (p = 0.040) (p = 0.005); however, the temporal quadrant was not reduced significantly, compared to the control value (p=0.907). Moreover, the only variable showing significant relation with RNFLT was duration of bipolar disorder (p=0.040). Conclusion Reduction of peripapilary RNFLT occurs in patients with bipolar disorder, and is related to the duration of disease. RNFLT can be a beneficial value for studying neurodegenerative changes over time towards detecting the severity and duration of disorder. Keywords Bipolar disorder . Optical coherence tomography . Retinal nerve fiber layer thickness
Introduction Bipolar disorder, a chronic mental disorder with young age onset, causes disability costing society a fortune [1]. Type one of bipolar disorder emerging with at least one episode of mania shows a prevalence of 1 % in general population. Several studies have been conducted on biological changes in mental disorders including bipolar disorder [2]. Despite ongoing researches, evidence of an association between genes and diagnostic criteria for bipolar disorder is still under the shadow of doubt. Intermediate phenotypes are objective measures of alteration in anatomy or function of the
Graefes Arch Clin Exp Ophthalmol
brain connecting genes to clinical manifestations, for example oculomotor abnormalities in schizophrenic patients or structural changes in the brain [3]. Magnetic resonance imaging (MRI) reveals structural brain pathologies including lateral ventrical enlargement, increased high density lesions in deep brain, and increasing thickness of the brain gray matter in lithium users [4]. These structural brain abnormalities present even in the first episode of bipolar disorder [5]. Some changes are progressive, such as reduction of brain gray matter in hipppocampus, fusiform gyrus, and cerebella and also in the temporal lobe correlating with the number of mood episodes [6]. Total brain gray matter volume is associated with the duration of bipolar disorder, and its changes can be interpreted as the evidence for the neurodegeneration process [7]. Gray matter volume is specifically used as an intermediate phenotype for schizophrenia [8–10]. There is genetic variability association with gray matter deficits in bipolar disorder [10], and also healthy bipolar offspring showing gray matter abnormalities [11]. Optical coherence tomography (OCT) is a noninvasive imaging technique providing high-resolution, cross-sectional images of the retina, and automatic measurement of retinal thickness and retinal nerve fiber layer thickness (RNFLT). Because retina contains axons and glia without myelin and originated from central nervous system (CNS), [12] RNFLT is valued as a more sensitive and also non-invasive parameter for evaluating neurodegeneration [13] in neurological disorders, particularly multiple sclerosis [14]. OCT which was initially used for assessment of glaucoma and macular disorders [15], is currently applied in neurological disorders including multiple sclerosis [12, 16–19], Alzheimer [20], and Parkinson [21]. In multiple sclerosis, cognitive and physical disability has been correlated with the degree of RNFL atrophy [22, 23]. Decreasing RNFLT has been reported in schizophrenia, probably relating to the neurodegeneration process [24, 25]. Duration of schizophrenia has been correlated to RNFL thinning, macular thinning, and reduction of macular volume [26], and severity of its symptoms to smaller macular volume [27]. Herein, due to atrophy of gray matter in patients with bipolar disorder, we studied the changes of peripapillary RNFLT by OCT as a possible marker for neurodegeneration comparing to normal subjects.
Methods In this prospective comparative case series, we compared 60 eyes of 30 patients with bipolar disorder to 60 eyes of 30 agematched healthy control subjects. Cases included inpatients and outpatients from the psychiatry department at Imam Hossein Hospital, and the controls were healthy individuals accompanying the patients in the outpatient ophthalmology
department of the same hospital. Patients were not masked to the diagnosis. The study was approved by the Ethics Committee of the Ophthalmic Research Center at Shahid Beheshti University of Medical Sciences, and followed the tenets of the Declaration of Helsinki. After explaining the purpose of the study, written informed consent was obtained from all subjects. The patients underwent a complete psychiatric examination including a semistructured clinical interview following the verified Persian translation SCID outline for bipolar disorder diagnosis and fulfilling the inclusion and exclusion DSM-IV-TR criteria of bipolar disorder [28, 29]. Moreover, for both the cases and controls a complete ophthalmologic examination, including assessment of best-corrected visual acuity (BCVA), intraocular pressure (IOP) measurement, slit-lamp biomicroscopy, and dilated fundus examination was done. Only participants with BCVA of 20/20 or better with normal IOP were included. An exclusion criteria was posterior pole pathology, including optic nerve and retinal pathology and any media haziness causing an interruption in ocular and OCT examination. Other exclusion criteria included psychiatric and neurological diseases with gray matter defects leading to RNFLT thinning such as multiple sclerosis, dementia, substance abuse [30], and also in order to reduce interfering factors, such as obsessive compulsive disorder. Peripapillary RNFLT was measured in all eyes on a circle at 3.4 mm around the optic disc in four superior, temporal, inferior, and nasal quadrants using OCT (3D OCT-1000, Topcon Corporation, Tokyo, Japan) by an experienced optometrist who was masked to the patient group. Scans with a quality factor of less than 60 and blinking during the scanning process were excluded. The thickness of each quadrant and the mean of all were used for analysis. Since all previous studies on RNFLT in psychiatric disorders were conducted on schizophrenia [24, 25, 27], emphasizing gray matter loss in psychosis whether affective or not [9], both current psychosis and history of it were considered in present analysis. Variables such as age of onset, duration, smoking, psychosis, mania and depression episodes in the case group and their relations with RNFLT were assessed by OCT. To evaluate the assumption of normality in the data, the Kolmogorov–Smirnov test was used; and to describe the data, frequency (percent), mean±SD, median, and range were used. For evaluation of the differences between the two groups, we utilized t-test, Mann–Whitney, and chi-square tests; and to compare groups while considering the correlation of eyes in one subject, generalized estimating equation (GEE) analysis was used. P-value less than 0.05 was considered as statistically significant. All statistical analysis were performed by SPSS software (Version 21.0, Microsoft Co., Chicago, IL, USA).
Graefes Arch Clin Exp Ophthalmol Table 1
Demographic specifications of participants in both groups
Parameter
Total
Group
P
Case
Control
Age (years)
Mean±SD Median (range)
32.5±9.3 31 (15 to 55)
33.8±9.2 32.5 (20 to 55)
31.2±9.5 30.5 (15 to 53)
0.285a
Sex
Male Female No Yes Mean±SD Median (range) Mean±SD Median (range) Mean±SD Median (range) Mean±SD Median (range)
48 (80.0 %) 12 (20.0 %) 35 (58.3 %) 25 (41.7 %) 14.7±9.5 12 (2 to 40) 8.7±7 8 (0.2 to 24) 0.01±0.02 0 (0 to 0.1) 13.6±1.7 14 (10 to 16)
24 (80.0 %) 6 (20.0 %) 12 (40.0 %) 18 (60.0 %) 16.3±10.5 16.5 (2 to 40) 10.2±7.4 10 (0.2 to 24) 0.01±0.02 0 (0 to 0.1) 13.7±1.9 14 (10 to 16)
24 (80.0 %) 6 (20.0 %) 23 (76.7 %) 7 (23.3 %) 10.4±4 10 (5 to 15) 4.7±3.4 5 (1 to 10) 0.01±0.02 0 (0 to 0.1) 13.4±1.5 13 (11 to 16)
1d
Smoking Smoking packs/ year Smoking duration (years) BCVA IOP (mmHg)
0.004d 0.185b 0.093b 0.449c 0.499c
BCVA Best-corrected visual acuity, IOP Intraocular pressure, SD Standard deviation a b
Based on t-test Based on Mann–Whitney test
c
Based on GEE analysis
d
Based on chi-square test
Results Mean age of bipolar patients and controls was 32.5±9.3 and 31.2±9.5 years respectively. Eighty percent of both groups were males. The number of smokers in the case group was 60.0 %, significantly higher than among control subjects (23.7 %). Cases had 10 years of smoking versus 5 years in controls, and 16.5 versus ten for number of cigarettes per day. Mean IOP was 13.6±1.7 and 13.7±1.9 mmHg for patients and control subjects respectively. Comparing the characteristic of cases and controls, the only meaningful difference was in the percentage of smokers (p value=0.004) (Table 1). There was no significant correlation between mean OCT thickness and either depression episode number or manic episode number (p=0. 960 and p=0. 627) and also no significant relationship between mean OCT thickness and age of onset (p=0. 421). Furthermore, no significant relationship was found between mean OCT thickness and smoking in the two groups (p=0. 728 and p=0. 424 in the case and control groups respectively). This correlation was not seen with the number of cigarettes (p=0. 794 and p= 0.191 in case and control groups respectively), nor with smoking duration (p=0.934 and p=0.546 in case and control groups respectively). Age of onset in the patients group was 23.2±5.4 years. Previous occurrence of psychosis was 66.7 %, and the majority of patients (83.3 %) were psychotic at the time of OCT scan. A total of 86.7 % of patients had only one episode of mania; a significant proportion of them (46.7 %) had no
episode of depression, and 36.7 % had only one episode. Among these variables, only duration of disease showed a significant relationship to RNFLT (Table 2). Moreover, we assessed the relationship of any kind of disorder (mania or depression) at presentation with RNFLT, and did not find any significant correlation (p=0.399). Mean RNFL thickness was 99±8 and 106±8 mμ in the case and control groups respectively, showing a significant difference (p=0.001). Mean RNFLT of inferior, superior, and nasal quadrants in bipolar patients was significantly less than that of corresponding quadrants in controls (p
NEUROPHTHALMOLOGY
Peripapillary retinal nerve fiber layer thickness in bipolar disorder Ali Mehraban 1 & Seyed Mehdi Samimi 1 & Morteza Entezari 2 & Mohammad Hassan Seifi 2 & Maryam Nazari 3 & Mehdi Yaseri 4
Received: 13 June 2014 / Revised: 12 December 2014 / Accepted: 2 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Background To compare peripapillary retinal nerve fiber layer thickness (RNFLT) between patients with bipolar disorder and a control group by optical coherence tomography (OCT). Methods This prospective comparative case series included 60 eyes of 30 patients with bipolar disorder and 60 eyes of The work was carried out by Ophthalmic Research Centre Shahid Beheshti Medical University. Ali Mehraban is a psychiatrist, Psychiatry Research Center, Department of Psychiatry, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Seyed Mehdi Samimi is an associate professor of Psychiatry, Psychiatry Research Center, Department of Psychiatry, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Morteza Entezari is an associate professor of Ophthalmology, Ophthalmic Research Center, Department of Ophthalmology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Mohammad Hassan Seifi is an ophthalmologist, Ophthalmic Research Center, Department of Ophthalmology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Maryam Nazary is a neurologist, Department of Neurology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, and holds a MD. Mehdi Yaseri is a biostatistician, Department of Epidemiology and Biostatistics, Tehran University of Medical Sciences, and holds a PhD. * Morteza Entezari [email protected] 1
Psychiatry Research Center, Department of Psychiatry, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran
2
Ophthalmic Research Center, Department of Ophthalmology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, Shahid Madani Ave., 1617763141 Tehran, Iran
3
Department of Neurology, Imam Hossein Medical Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran
4
Department of Epidemiology and Biostatistics, Tehran University of Medical Sciences, Tehran, Iran
30 age-matched healthy control subjects. Using OCT, peripapillary RNFLT of the 4 quadrants and the mean of them was compared between the two groups. Variables such as age of onset, duration, smoking, psychosis, mania and depression episodes in the case group and their relationships with RNFLT were evaluated by OCT. Results Mean RNFLT was 99±8 in the case group, significantly less than the106±8 mμ in the control group (p=0.001). The inferior, superior, and nasal quadrants in the case and control groups showed significant difference in RNFLT (p < 0.001) (p = 0.040) (p = 0.005); however, the temporal quadrant was not reduced significantly, compared to the control value (p=0.907). Moreover, the only variable showing significant relation with RNFLT was duration of bipolar disorder (p=0.040). Conclusion Reduction of peripapilary RNFLT occurs in patients with bipolar disorder, and is related to the duration of disease. RNFLT can be a beneficial value for studying neurodegenerative changes over time towards detecting the severity and duration of disorder. Keywords Bipolar disorder . Optical coherence tomography . Retinal nerve fiber layer thickness
Introduction Bipolar disorder, a chronic mental disorder with young age onset, causes disability costing society a fortune [1]. Type one of bipolar disorder emerging with at least one episode of mania shows a prevalence of 1 % in general population. Several studies have been conducted on biological changes in mental disorders including bipolar disorder [2]. Despite ongoing researches, evidence of an association between genes and diagnostic criteria for bipolar disorder is still under the shadow of doubt. Intermediate phenotypes are objective measures of alteration in anatomy or function of the
Graefes Arch Clin Exp Ophthalmol
brain connecting genes to clinical manifestations, for example oculomotor abnormalities in schizophrenic patients or structural changes in the brain [3]. Magnetic resonance imaging (MRI) reveals structural brain pathologies including lateral ventrical enlargement, increased high density lesions in deep brain, and increasing thickness of the brain gray matter in lithium users [4]. These structural brain abnormalities present even in the first episode of bipolar disorder [5]. Some changes are progressive, such as reduction of brain gray matter in hipppocampus, fusiform gyrus, and cerebella and also in the temporal lobe correlating with the number of mood episodes [6]. Total brain gray matter volume is associated with the duration of bipolar disorder, and its changes can be interpreted as the evidence for the neurodegeneration process [7]. Gray matter volume is specifically used as an intermediate phenotype for schizophrenia [8–10]. There is genetic variability association with gray matter deficits in bipolar disorder [10], and also healthy bipolar offspring showing gray matter abnormalities [11]. Optical coherence tomography (OCT) is a noninvasive imaging technique providing high-resolution, cross-sectional images of the retina, and automatic measurement of retinal thickness and retinal nerve fiber layer thickness (RNFLT). Because retina contains axons and glia without myelin and originated from central nervous system (CNS), [12] RNFLT is valued as a more sensitive and also non-invasive parameter for evaluating neurodegeneration [13] in neurological disorders, particularly multiple sclerosis [14]. OCT which was initially used for assessment of glaucoma and macular disorders [15], is currently applied in neurological disorders including multiple sclerosis [12, 16–19], Alzheimer [20], and Parkinson [21]. In multiple sclerosis, cognitive and physical disability has been correlated with the degree of RNFL atrophy [22, 23]. Decreasing RNFLT has been reported in schizophrenia, probably relating to the neurodegeneration process [24, 25]. Duration of schizophrenia has been correlated to RNFL thinning, macular thinning, and reduction of macular volume [26], and severity of its symptoms to smaller macular volume [27]. Herein, due to atrophy of gray matter in patients with bipolar disorder, we studied the changes of peripapillary RNFLT by OCT as a possible marker for neurodegeneration comparing to normal subjects.
Methods In this prospective comparative case series, we compared 60 eyes of 30 patients with bipolar disorder to 60 eyes of 30 agematched healthy control subjects. Cases included inpatients and outpatients from the psychiatry department at Imam Hossein Hospital, and the controls were healthy individuals accompanying the patients in the outpatient ophthalmology
department of the same hospital. Patients were not masked to the diagnosis. The study was approved by the Ethics Committee of the Ophthalmic Research Center at Shahid Beheshti University of Medical Sciences, and followed the tenets of the Declaration of Helsinki. After explaining the purpose of the study, written informed consent was obtained from all subjects. The patients underwent a complete psychiatric examination including a semistructured clinical interview following the verified Persian translation SCID outline for bipolar disorder diagnosis and fulfilling the inclusion and exclusion DSM-IV-TR criteria of bipolar disorder [28, 29]. Moreover, for both the cases and controls a complete ophthalmologic examination, including assessment of best-corrected visual acuity (BCVA), intraocular pressure (IOP) measurement, slit-lamp biomicroscopy, and dilated fundus examination was done. Only participants with BCVA of 20/20 or better with normal IOP were included. An exclusion criteria was posterior pole pathology, including optic nerve and retinal pathology and any media haziness causing an interruption in ocular and OCT examination. Other exclusion criteria included psychiatric and neurological diseases with gray matter defects leading to RNFLT thinning such as multiple sclerosis, dementia, substance abuse [30], and also in order to reduce interfering factors, such as obsessive compulsive disorder. Peripapillary RNFLT was measured in all eyes on a circle at 3.4 mm around the optic disc in four superior, temporal, inferior, and nasal quadrants using OCT (3D OCT-1000, Topcon Corporation, Tokyo, Japan) by an experienced optometrist who was masked to the patient group. Scans with a quality factor of less than 60 and blinking during the scanning process were excluded. The thickness of each quadrant and the mean of all were used for analysis. Since all previous studies on RNFLT in psychiatric disorders were conducted on schizophrenia [24, 25, 27], emphasizing gray matter loss in psychosis whether affective or not [9], both current psychosis and history of it were considered in present analysis. Variables such as age of onset, duration, smoking, psychosis, mania and depression episodes in the case group and their relations with RNFLT were assessed by OCT. To evaluate the assumption of normality in the data, the Kolmogorov–Smirnov test was used; and to describe the data, frequency (percent), mean±SD, median, and range were used. For evaluation of the differences between the two groups, we utilized t-test, Mann–Whitney, and chi-square tests; and to compare groups while considering the correlation of eyes in one subject, generalized estimating equation (GEE) analysis was used. P-value less than 0.05 was considered as statistically significant. All statistical analysis were performed by SPSS software (Version 21.0, Microsoft Co., Chicago, IL, USA).
Graefes Arch Clin Exp Ophthalmol Table 1
Demographic specifications of participants in both groups
Parameter
Total
Group
P
Case
Control
Age (years)
Mean±SD Median (range)
32.5±9.3 31 (15 to 55)
33.8±9.2 32.5 (20 to 55)
31.2±9.5 30.5 (15 to 53)
0.285a
Sex
Male Female No Yes Mean±SD Median (range) Mean±SD Median (range) Mean±SD Median (range) Mean±SD Median (range)
48 (80.0 %) 12 (20.0 %) 35 (58.3 %) 25 (41.7 %) 14.7±9.5 12 (2 to 40) 8.7±7 8 (0.2 to 24) 0.01±0.02 0 (0 to 0.1) 13.6±1.7 14 (10 to 16)
24 (80.0 %) 6 (20.0 %) 12 (40.0 %) 18 (60.0 %) 16.3±10.5 16.5 (2 to 40) 10.2±7.4 10 (0.2 to 24) 0.01±0.02 0 (0 to 0.1) 13.7±1.9 14 (10 to 16)
24 (80.0 %) 6 (20.0 %) 23 (76.7 %) 7 (23.3 %) 10.4±4 10 (5 to 15) 4.7±3.4 5 (1 to 10) 0.01±0.02 0 (0 to 0.1) 13.4±1.5 13 (11 to 16)
1d
Smoking Smoking packs/ year Smoking duration (years) BCVA IOP (mmHg)
0.004d 0.185b 0.093b 0.449c 0.499c
BCVA Best-corrected visual acuity, IOP Intraocular pressure, SD Standard deviation a b
Based on t-test Based on Mann–Whitney test
c
Based on GEE analysis
d
Based on chi-square test
Results Mean age of bipolar patients and controls was 32.5±9.3 and 31.2±9.5 years respectively. Eighty percent of both groups were males. The number of smokers in the case group was 60.0 %, significantly higher than among control subjects (23.7 %). Cases had 10 years of smoking versus 5 years in controls, and 16.5 versus ten for number of cigarettes per day. Mean IOP was 13.6±1.7 and 13.7±1.9 mmHg for patients and control subjects respectively. Comparing the characteristic of cases and controls, the only meaningful difference was in the percentage of smokers (p value=0.004) (Table 1). There was no significant correlation between mean OCT thickness and either depression episode number or manic episode number (p=0. 960 and p=0. 627) and also no significant relationship between mean OCT thickness and age of onset (p=0. 421). Furthermore, no significant relationship was found between mean OCT thickness and smoking in the two groups (p=0. 728 and p=0. 424 in the case and control groups respectively). This correlation was not seen with the number of cigarettes (p=0. 794 and p= 0.191 in case and control groups respectively), nor with smoking duration (p=0.934 and p=0.546 in case and control groups respectively). Age of onset in the patients group was 23.2±5.4 years. Previous occurrence of psychosis was 66.7 %, and the majority of patients (83.3 %) were psychotic at the time of OCT scan. A total of 86.7 % of patients had only one episode of mania; a significant proportion of them (46.7 %) had no
episode of depression, and 36.7 % had only one episode. Among these variables, only duration of disease showed a significant relationship to RNFLT (Table 2). Moreover, we assessed the relationship of any kind of disorder (mania or depression) at presentation with RNFLT, and did not find any significant correlation (p=0.399). Mean RNFL thickness was 99±8 and 106±8 mμ in the case and control groups respectively, showing a significant difference (p=0.001). Mean RNFLT of inferior, superior, and nasal quadrants in bipolar patients was significantly less than that of corresponding quadrants in controls (p
