Social function in schizophrenia and schizoaffective disorder: Associations with personality, symptoms and neurocognition
© Lysaker and Davis; licensee BioMed Central Ltd. 2004
Received: 22 December 2003
Accepted: 16 March 2004
Published: 16 March 2004
Research has indicated that stable individual differences in personality exist among persons with schizophrenia spectrum disorders predating illness onset that are linked to symptoms and self appraised quality of life. Less is known about how closely individual differences in personality are uniquely related to levels of social relationships, a domain of dysfunction in schizophrenia more often linked in the literature with symptoms and neurocognitive deficits. This study tested the hypothesis that trait levels of personality as defined using the five-factor model of personality would be linked to social function in schizophrenia.
A self-report measure of the five factor model of personality was gathered along with ratings of social function, symptoms and assessments of neurocognition for 65 participants with schizophrenia or schizoaffective disorder.
Univariate correlations and stepwise multiple regression indicated that frequency of social interaction was predicted by higher levels of the trait of Agreeableness, fewer negative symptoms, better verbal memory and at the trend level, lesser Neuroticism (R2 = .42, p < .0001). In contrast, capacity for intimacy was predicted by fewer negative symptoms, higher levels of Agreeableness, Openness, and Conscientiousness and at the trend level, fewer positive symptoms (R2 = .67, p < .0001).
Taken together, the findings of this study suggest that person-centered variables such as personality, may account for some of the broad differences seen in outcome in schizophrenia spectrum disorders, including social outcomes. One interpretation of the results of this study is that differences in personality combine with symptoms and neurocognitive deficits to affect how persons with schizophrenia are able to form and sustain social connections with others.
KeywordsSchizophrenia Personality Symptoms Coping Quality of life
Interest has increasingly grown in understanding how differences in personality may affect outcome in schizophrenia [1, 2]. Just as in a wide range of other severe and debilitating medical conditions [3–7], the manner in which people interpret and respond to a life touched by schizophrenia may deeply impact upon the recovery process [8–11].
To date, one model of personality that has shown some promise in helping to systematically document the types of individual differences that help or hinder outcome in schizophrenia, is the "Five factor" model . This model posits five endogenous traits  along which all persons vary, regardless of their socioeconomic status or culture and which exert an enduring impact on behavior, affect and cognition across the lifespan . These five dimensions are Neuroticism, or vulnerability to emotional instability and self-consciousness, Extraversion, or the tendency to be warm and outgoing; Openness, or the cognitive disposition to creativity and aesthetics; Agreeableness, or the tendency to be comfortable with social interactions, and Conscientiousness, or the tendency towards dutifulness and competence [12, 15]. Each of these dimensions is conceptualized as a "basic tendency" which interacts with external influences to shape how persons adapt and form their self-concept.
Beyond its intuitive appeal as a model for understanding individual differences in schizophrenia, research has suggested that the traits of the five factor model can be detected in schizophrenia  and that, as in the general population, these traits are relatively stable over time . Additionally, persons with schizophrenia tend to present with a different pattern of these traits, endorsing higher levels of Neuroticism and lower levels of Extraversion, Openness, Agreeableness and Conscientiousness than community controls [16, 18]. Regarding clinical outcomes, levels of neuroticism and agreeableness have been linked to heightened symptom levels [18, 19]. Neuroticism, Extraversion and Agreeableness have also been linked to poorer life satisfaction  and to more avoidant coping [18, 21]. Other assessments of neuroticism and extraversion from slightly different trait models have suggested both are related to symptoms  and work function  and may predate the onset of symptoms potentially reflecting risk factors for the development of schizophrenia spectrum disorders [24, 25]. Lastly, other assessments using competing models of temperament and character have also found personality variables linked with patterns of substance abuse  and lesser levels of quality of life [27–29].
While this literature points to a link between personality and clinical outcome, curiously less has been studied about impact of the traits of the five-factor model on the ability of persons with schizophrenia to form and sustain close interpersonal relationships. While research linking poor social function to negative symptoms and neurocognitive impairments [30–33] has generated considerable excitement, the influence of individual differences has been somewhat neglected. Might not higher levels of Neuroticism as well as lower levels of Extraversion, Openness, Agreeableness and Conscientiousness also uniquely contribute to poorer level of intimate connections to others and one's community, along with levels of neurocognitive symptoms and neurocognitive deficits?
To investigate this possibility we assessed interpersonal and community function using the Quality of Life Scale [; QOLS] and the five traits of the five factor model using the NEO . Concurrently we also assessed positive and negative domains of psychopathology with the Positive And Negative Syndrome Scale [; PANSS] and three aspects of neurocognitive function linked to community function: verbal memory, executive function and premorbid intellectual function. Three primary predictions were made: higher levels of Neuroticism and lower levels of Extraversion, Openness, Agreeableness and Conscientiousness would uniquely contribute to lesser amount of social contact, fewer of the resources needed for intimacy and poorer community function, each as assessed on the QOLS. It was also predicted that these associations would exist semi-independently of the effects of negative symptoms and neurocognitive impairments.
Sixty-five males with SCID  confirmed DSM IV diagnoses of schizophrenia or schizoaffective disorder were recruited from a comprehensive day hospital at a VA Medical Center. All participants were receiving ongoing outpatient treatment and were in a post-acute or stable phase of their disorder. Clinical stability was defined as no hospitalizations or changes in medication or housing in the last month. Participants with organic brain syndrome or history of mental retardation documented in a chart review were also excluded. Participants had a mean age of 47.5 (sd = 9), a mean educational level of 12.2 (sd = 1.7) and a mean of 8 (sd = 7.7) lifetime hospitalizations with the first on average occurring at the age of 25 (sd = 5.8). Thirty-six were Caucasians, 28 African-Americans, and one Latino. Forty-one had schizophrenia and 24 schizoaffective disorder. Forty were being prescribed atypical anti-psychotic medication at baseline, 16 a combination of typical and atypical, and 9 typical anti-psychotic medication. The mean chlorpromazine equivalence dose was 860 mg (sd = 1010).
Positive And Negative Syndrome Scale: [ PANSS; Kay et al, 1987] is a 30-item rating scale completed by clinically trained research staff at the conclusion of chart review and a semi-structured interview. It is one of the most widely used semi-structured interviews for assessing the wide range of psychopathology in schizophrenia. For the purposes of this study three of the PANSS factor analytically-derived components scores were utilized: Positive, Negative, and Emotional Discomfort . The other components scores not used here are Cognitive and Excitement symptoms. The five-factor structure of the PANSS has been replicated several times . Assessment of inter-rater reliability for this study found good to excellent with intraclass correlations ranging from .84 to .93.
NEO Five-Factor Inventory (form s):  is a self-report assessment of personality dimensions based on the five-factor model of personality. This test presents participants with 60 statements that they are asked to rate on a likert scale as describing or not describing their attitudes and behavior. The NEO form s generates percentile scores for the personality dimensions of Neuroticism, Extraversion, Openness, Agreeableness and Conscientiousness. The short form of the NEO has been used successfully in other studies of personality and schizophrenia [16–18]. For the purposes of this study we examined the Neuroticism, Extraversion, Openness, Agreeableness and Conscientiousness scores.
Hopkins Verbal Learning Test [; HVLT] is an auditory verbal memory test designed to measure verbal memory and learning potential. In this test the experimenter verbally presents a list of words each belonging to one of several semantic categories three times and then after a delay asks the participant how many words they can recall. For the purposes of this study we utilized the age corrected t score for recall after the delay.
Wisconsin Card Sorting Test [; WCST] is a neuropsychological test sensitive to impairments in executive function. It asks participants to sort cards that vary according to an unarticulated matching principle that changes after a certain number of correct responses. The current study utilized the age and education corrected t score for perseverative errors. This score is of particular interest since it is hypothesized as most closely relating to inflexibility of abstract reasoning.
The Vocabulary subtest of the WAIS-III  assesses participants' knowledge of vocabulary. This subtest has been widely used as a brief assessment of general verbal intellectual function.
Quality of Life Scale [ QOLS] is a 21-item scale completed by clinically trained research staff following a semi-structured interview and chart review. For the purposes of this study, we were interested in three of the four factor scores of the QOL. The first, "Interpersonal Relations," measures the frequency of recent social contacts and includes separate assessments, for example, of frequency of contacts with friends and acquaintances. The second, "Intrapsychic Foundations," measures qualitative aspects of interpersonal relationships and includes assessments, for example, of empathy for others. The third, "Common Objects and Activities," assesses community involvement in terms of participation in common activities and possession of common objects that denote such participation. The fourth, "Instrumental Role," was not of interest, as this scale taps vocational function and all participants were entering vocational rehabilitation because they were unemployed and thus there was no variation in this scale. Good to excellent inter-rater reliability was found for the three QOL factor scores for this study, with intraclass correlations for blind raters observing the same interview ranging from .85 to .93. Although originally created to assess negative symptoms in schizophrenia the QOLS has been widely used to study social function among persons with schizophrenia .
Following informed consent diagnoses were determined using the Structured Clinical Interview for DSM IV  conducted by a clinical psychologist (PL). Following the SCID, participants were administered the PANSS and QOLS interviews, NEO and neurocognitive testing. PANSS and QOLS ratings were performed blind to responses to the NEO and neurocognitive test scores. Neurocognitive testing, QOLS and PANSS interviews were conducted by trained research assistants with a minimum of a B.A. degree in a field related to psychology.
Mean NEO percentile scores were: Neuroticism M = 61.8 (SD = 9.9), Extraversion M = 44.4 (SD = 9.8), Openness M = 45.5 (SD = 8.3), Agreeableness M = 44.6 (SD = 11.2), and Conscientiousness M = 44.4 (SD = 9.8). Mean PANSS components scores were: Positive M = 17.3 (SD = 5.5), and Negative M = 20.4 (SD = 5.1). Mean neurocognitive testing scores were: HVLT delayed recall T score: M = 34.3 (SD = 10.6) WCST Perseverative errors T score M = 38.4 (SD = 12.2) and Vocabulary subtest: M = 7.5 (SD = 2.8). Correlations of NEO scores with PANSS and neurocognitive test scores revealed Positive symptoms were related to Neuroticism (r = .28, p < .05) and Agreeableness (r = -.49, p < .001). Openness was related to Negative symptoms (r = -.29 p < .05) HVLT (r = .38, p < .01) and Vocabulary subtest (r = .32, p < .05). Extraversion was related to WCST (r = -.27, p < .05) and Conscientiousness was related to Vocabulary (r = -.28, P < .05). The NEO, PANSS and QOLS scores of participants with schizoaffective disorder did not differ significantly from those of participants with schizophrenia.
Personality, symptom and neurocognitive correlates of three dimensions of social function (n = 65)
Quality of life subscales
Common objects and activities
HVLT Delayed Recall
WCST Perseverative Errors
WAIS III Vocabulary
PANSS Positive symptoms
PANSS Negative symptoms
Multiple regressions predicting QOL scores from NEO, PANSS, and neurocognitive test scores
Measure of social function
Contributing PANSS and NEO components
QOL Interpersonal relationships
PANSS Negative symptoms
QOL Intrapsychic foundations
PANSS Negative symptoms
PANSS Positive symptoms
Results of this study are consistent with previous studies linking personality with general outcome including sense of well being in schizophrenia spectrum disorders. In particular, participants with more social ties tended to have lesser levels of Neuroticism, and higher levels of Agreeableness. Participants with greater capacities for intimacy similarly tended to have lesser levels of Neuroticism, higher levels of Openness, Agreeableness and Conscientiousness. Replicating previous studies, better verbal memory and premorbid intellectual function and fewer Positive and Negative Symptoms also predicted more social ties, while better verbal memory and fewer Positive and Negative Symptoms predicted a greater capacity for intimacy.
Given the complex interrelationships among personality, symptoms, neurocognition and social function, it is even more striking that when entered into a regression, personality variables tended to capture unique and significant proportions of the variances, despite levels of negative symptoms and in the case of interpersonal relations, verbal memory also capturing unique portions of the variance. Also notable, was that taken together, personality, symptoms and neurocognition were able to account for between two fifths and two thirds of the variance in two of the three QOLS measures.
Explanations for heterogeneity within schizophrenia have included issues ranging from differences in pathophysiology [44, 45] to prevailing social conditions [46, 47]. Taken together, the findings of this study suggest that person centered variables such as personality, may also account for some of the broad differences seen in outcome in this disorder, including social outcomes. One intuitively appealing interpretation of this data is that differences in personality combine with symptoms and neurocognitive deficits to affect how persons with schizophrenia are able to form and sustain social connections with others. Of note, the correlational nature of this study precludes drawing any firm conclusions about causality and thus it may be that other factors not measured, such as stigma, account for the observed relationships between personality and social function. It is also possible that the experience of social rejection affects personality as measured using the NEO.
While the general hypotheses regarding personality and social function were confirmed, surprisingly, Extraversion did not seem to be related to any QOLS measure. This may suggest that Extraversion is not particularly advantageous to persons with schizophrenia, at least in terms of sociability. As we have hypothesized elsewhere [22, 23] perhaps being socially outgoing when one has numerous deficits and idiosyncratic views may make one a target for stigmatization and rejection, thus negating perhaps any social gains. Clearly, however, this is speculation at present and future studies are needed to examine this question. Also, community function was related to neurocognition alone and no relationships were found with personality. This may suggest that participation in community is more greatly mediated by biological factors and social factors not assessed here.
Lastly, there are several other methodological limitations to this study. Generalization of findings is limited by sample composition. Participants were almost exclusively male and in their 40's who were involved in treatment. It may be that a different relationship exists between personality and social function among females or among younger males with schizophrenia, or persons who decline treatment. The battery assessing neurocognition was also limited in size and scope. Thus more research is necessary with broader samples and instrumentation. In particular, more "fine-grained" assessments of function as well as longitudinal assessments of personality, behavior and psychopathology may find associations between behavior and personality that have important implications for treatment and rehabilitation. For instance, personality may prove to be an easily measurable personal characteristic that predicts outcome. Thus it may prove efficacious to identify subgroups of persons who may receive special benefit from interventions that emphasize identifying and coping with painful affects [e.g. [2, 48]], or help to manage chronically unstable emotional states.
- Davidson L, Strauss J: Sense of self in recovery from mental illness. Brit J Med Psychol 1992, 65: 131–145.PubMedView ArticleGoogle Scholar
- Lysaker PH, France C: Psychotherapy as an element in supportive employment for persons with severe and persistent mental illness. Psychiatry 1999, 62: 209–222.PubMedGoogle Scholar
- Kleinman A: The illness narratives: Suffering, healing and the human condition. Basic Books: New York 1988.Google Scholar
- Mechanic D: Sociological dimensions of illness behavior. Social Science Medicine 1995, 41: 1207–1216. 10.1016/0277-9536(95)00025-3PubMedView ArticleGoogle Scholar
- Novy DM, Nelson DV, Fancis DJ, Turk DC: Perspectives of chronic pain: An evaluative comparison of restrictive and comprehensive models. Psychol Bull 1995, 118: 238–247. 10.1037//0033-2909.118.2.238PubMedView ArticleGoogle Scholar
- Schreurs KMG, de Ridder DTD: Integration of coping and social support perspectives: Implications for the study of adaptation to chronic diseases. Clin Psychol Rev 1997, 17: 89–112. 10.1016/S0272-7358(96)00050-5PubMedView ArticleGoogle Scholar
- Schwartz CE, Rogers M: Designing a psychosocial intervention to teach coping flexibility. Rehabilitation Psychology 1994, 39: 57–72. 10.1037//0090-55184.108.40.206View ArticleGoogle Scholar
- Lysaker PH, Lysaker JT: Narrative Structure in Psychosis: Schizophrenia and Disruptions in the Dialogical Self. Theory and Psychology 2002, 12: 207–220. 10.1177/0959354302012002630View ArticleGoogle Scholar
- Roe D: A perspective study on the relationship between self-esteem and functioning during the first year being hospitalized for psychosis. J Nerv Ment Dis 2003,191(1):45–49. 10.1097/00005053-200301000-00008PubMedView ArticleGoogle Scholar
- Thompson Jr EH: Variation in the self-concept of young adult chronic patients: Chronicity reconsidered. Hosp Community Psychiatry 1988, 39: 771–775.Google Scholar
- Warner R, Taylor D, Powers M, Hyman R: Acceptance of the mental illness label by psychotic patients: effects on functioning. Am J Orthopsychiatry 1989, 59: 398–409.PubMedView ArticleGoogle Scholar
- McCrae RR, Costa JT: Personality trait structure as a human universal. Am Psychol 1997, 52: 509–516. 10.1037//0003-066X.52.5.509PubMedView ArticleGoogle Scholar
- Jang KL, Livesley WJ, Vernon PA: Heritability of the big five personality dimensions and their facets: a twin study. J Pers 1996, 64: 577–591.PubMedView ArticleGoogle Scholar
- Von Dras DD, Siegler IC: Stability in extraversion and aspects of social support at mid life. J Pers Soc Psychol 1997, 72: 233–241. 10.1037//0022-35220.127.116.11PubMedView ArticleGoogle Scholar
- McCrae RR, Costa JT: Toward a new generation of personality theories: Theoretical contexts for the five-factor model. In: The Five-Factor Model of Personality: Theoretical Perspectives (Edited by: Wiggins JS). NY, NY: Guilford Press 1996, 51–87.Google Scholar
- Gurrera RJ, Nestor PG, O'Donnel BF: Personality traits in schizophrenia: Comparison with a community sample. J Nerv Ment Dis 2000, 188: 31–35. 10.1097/00005053-200001000-00006PubMedView ArticleGoogle Scholar
- Kentros MD, Smith TE, Hull J, McKee M, Terkelsen K, Capalbo C: Stability of personality traits in schizophrenia and schizoaffective disorder: A pilot project. J Nerv Ment Dis 1997, 185: 549–555. 10.1097/00005053-199709000-00003PubMedView ArticleGoogle Scholar
- Lysaker PH, Wilt MA, Plascek-Hallberg CD, Brenner C, Clements CA: The association of personality with symptoms and coping in schizophrenia. J Nerv Ment Dis 2003, 191: 80–86. 10.1097/00005053-200302000-00003PubMedGoogle Scholar
- Lysaker PH, Lancaster RS, Nees M, Davis LW: Neuroticism and visual memory impairments as predictors of the severity of delusions in schizophrenia. Psychiatry Res 2003,119(3):87–292. 10.1016/S0165-1781(03)00169-0View ArticleGoogle Scholar
- Kentros MD, Terkelsen K, Hull J, Smith TE, Goodman M: The relationship between personality and and quality of life in persons with schizoaffective disorder and schizophrenia. Qual Life Res 1997, 6: 118–122. 10.1023/A:1026433932142PubMedView ArticleGoogle Scholar
- Lysaker PH, Bryson GJ, Marks KA, Greig T, Bell MD: Coping in Schizophrenia: Associations with Neurocognitive Deficits and Personality. Schizophr Bull, in press.Google Scholar
- Lysaker PH, Bell MD, Kaplan E, Greig TC, Bryson GJ: Personality and Psychopathology in Schizophrenia: The Association Between Personality Traits and Symptoms. Psychiatry 1999, 62: 36–48.PubMedGoogle Scholar
- Lysaker PH, Bell MD, Kaplan E, Bryson GJ: Personality and Psychosocial dysfunction in schizophrenia: The association of extraversion and neuroticism with deficits in work performance. Psychiatry Res 1998, 80: 61–68. 10.1016/S0165-1781(98)00049-3PubMedView ArticleGoogle Scholar
- Berenbaum H, Fujita F: Schizophrenia and personality: Exploring the boundaries and connections between vulnerability and outcome. J Abnorm Psychol 1994, 103: 148–158. 10.1037//0021-843X.103.2.241PubMedView ArticleGoogle Scholar
- van Os J, Jones PB: Neuroticism as a risk factor for schizophrenia. Psychol Med 2001, 31: 1129–1134. 10.1017/S0033291701004044PubMedView ArticleGoogle Scholar
- Van Ammers EC, Sellman D, Mulder RT: Temperament and substance abuse in schizophrenia. J Nerv Ment Dis 1997, 185: 283–288. 10.1097/00005053-199705000-00001PubMedView ArticleGoogle Scholar
- Ritsner M, Herman F, Gibel A: Satisfaction with quality of life varies with temperment types of patients with schizophrenia. J Nerv Ment Dis 2003, 191: 668–674. 10.1097/00005053-200305000-00003PubMedView ArticleGoogle Scholar
- Ecklund M: Personality and self variables: Important determinants of subjective quality of life in schizophrenia outpatients. Acta Psychiatr Scand 2003, 108: 134–144. 10.1034/j.1600-0447.2003.00103.xView ArticleGoogle Scholar
- Hansson L, Ecklund M, Bengtsson-Tops A: The relationship of personality dimensions as measured by the temperament and character inventory and quality of life in individuals with schizoaffective disorder living in the community. Qual Life Res 2001, 10: 133–139. 10.1023/A:1016790230228PubMedView ArticleGoogle Scholar
- Addington J, Addington D: Neurocognitive and social functioning in schizophrenia. Schizophr Bull 1999, 25: 173–182.PubMedView ArticleGoogle Scholar
- Carter M, Flesher S: The neurosociology of schizophrenia: Vulnerability and functional disability. Psychiatry 1995, 58: 209–221.PubMedGoogle Scholar
- Mueser KT, Bellack AS, Douglas M, Wade J: Prediction of social skill acquisition in schizophrenia and affective disorder patients with memory and symptomatology. Psychiatric Res 1992, 37: 281–296. 10.1016/0165-1781(91)90064-VView ArticleGoogle Scholar
- Green MF: What are the functional consequences of neurocognitive deficits in schizophrenia? Am J Psychiatry 1996,153(3):321–330.PubMedView ArticleGoogle Scholar
- Heinrichs DW, Hanlon TE, Carpenter WT: The Quality of Life Scale: An instrument for assessing the schizophrenic deficit syndrome. Schizophr Bull 1984, 10: 388–396.PubMedView ArticleGoogle Scholar
- Costa PT, McCrae RR: Revised NEO Personality Inventory and NEO Five factor Inventory. Odessa Fl: Psychological Assessment Resources 1992.Google Scholar
- Kay SR, Fizszbein A, Opler LA: The positive and negative syndrome scale for schizophrenia. Schizophr Bull 1987, 13: 261–276.PubMedView ArticleGoogle Scholar
- Spitzer R, Williams J, Gibbon M, First M: Structured Clinical Interview for DSM IV Biometrics Research: New York. 1994.Google Scholar
- Bell MD, Lysaker PH, Goulet JB, Milstein RM, Lindenmayer JP: Five-component model of schizophrenia: Assessing the factorial invariance of the PANSS. Psychiatry Res 1993, 52: 295–303. 10.1016/0165-1781(94)90075-2View ArticleGoogle Scholar
- Frederickson DH, Steiger JM, MacEwan GW, Altman S, Kopala LC, Flynn SW, Liddle PF, Honer WG: PANSS symptom factors in schizophrenia. Schizophr Res 1997, 24: 15. 10.1016/S0920-9964(97)82039-7View ArticleGoogle Scholar
- Brandt J: The Hopkins Verbal Learning Test: Development of a new memory test with six equivalent forms. Clin Neuropsychol 1991,5(2):125–142.View ArticleGoogle Scholar
- Heaton RK, Chelune GJ, Talley JL, Kay GG, Curtiss G: Wisconsin Card Sorting Test Manual: Revised and Expanded. Odessa: Psychological Assessment Resources, Inc 1993.Google Scholar
- Weschler D: Weschler Adult Intelligence Scale III. New York: Psychological Corporation 1997.Google Scholar
- Lysaker PH, Bell MD, Bryson GJ, Kaplan E: Psychosocial function and insight in schizophrenia. J Nerv Ment Dis 1998, 186: 432–436. 10.1097/00005053-199807000-00008PubMedView ArticleGoogle Scholar
- Buchanan RW, Carpenter WT: Domains of psychopathology: An approach to the reduction of heterogeneity in schizophrenia. J Nerv Ment Dis 1994, 182: 193–204.PubMedView ArticleGoogle Scholar
- Olney JW, Farber NB: Glutamate receptor dysfunction in schizophrenia. Arch Gen Psychiatry 1995, 52: 998–1007.PubMedView ArticleGoogle Scholar
- Allen JS: Are traditional societies schizophrenogenic? Schizophr Bull 1997, 23: 357–364.PubMedView ArticleGoogle Scholar
- Warner R: Recovery From Schizophrenia: Psychiatry and Political economy. New York: Routledge and Kegan Paul 1985.Google Scholar
- Hogarty GE, Kornblith SJ, Greenwald D, DiBarry AL, Cooley S, Ulrich RF, Carter M, Flesher S: Three year trials of personal therapy among schizophrenic patients living with or independent of family, I: Description of the study of effects on relapse rates. Am J Psychiatry 1997, 154: 1504–1513.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original URL.