ºÚÁÏÍø

Co-Morbid Psychopathology of Patients with Pathological Internet use and Alcoholism - A Comparative Study

te Wildt BT¹, Siebrasse P¹, Putzig I², Dillo W¹, Wiese B¹, Szycik GR¹, Ohlmeier MD³ and Wedegaertner F^1*

¹Department of Psychiatry, Social Psychiatry und Psychotherapy, Hanover Medical School (MHH), Carl-Neuberg-Straße 1, 30625 Hanover, Germany

²Department of Child and Adolescent Psychiatry and Psychotherapy, University Hospital Aachen, Pauwelstraße 30, 52074 Aachen, Germany

³Department of Psychiatry and Psychotherapy, Ludwig-Noll-Hospital, Klinikum Kassel, Dennhäuser Straße 156, 34134 Kassel, Germany

*Corresponding Author:

Felix Wedegaertner, MPH
Department of Psychiatry
Social Psychiatry and Psychotherapy
Hanover Medical School, 30625 Hannover, Germany
Tel: 0049-511-532-8526
Fax: 0049-511-532-8298
E-mail: wedegaertner.felix@mh-hannover.de

Received October 25, 2011; Accepted January 16, 2012; Published January 20, 2012

Citation:te Wildt BT, Siebrasse P, Putzig I, Dillo W, Wiese B, et al. (2012) Co- Morbid Psychopathology of Patients with Pathological Internet use and Alcoholism – A Comparative Study. J Addict Res Ther S6:002. doi:10.4172/2155-6105.S6-002

Copyright: © 2012 te Wildt BT, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Visit for more related articles at Journal of Addiction Research & Therapy

Abstract

Aims: There is increasing concern that the Internet and video games contain an addictive potential. However, it has been discussed, whether pathological Internet use (PIU) is to be diagnosed as an impulse control disorder as pathological gambling or in analogy to substance abuse disorders. Contributing to this discussion, the study compares psychopathological features and co-morbidities of patients with PIU and alcoholism. Methods: Both 25 Internet- and alcohol-dependent patients were assessed for psychopathological symptomatology and co-morbidity with the Structured Clinical Interview according to DSM-IV (SKID), the Symptom-Checklist (SCL-90R), Beck Depression Inventory (BDI), Connors’ Adult ADHD Rating Scales (CAARS) and the Barrat Impulsiveness Scale (BIS). For the PIU-patients the level of dependency was measured with the Internet Addiction Scale (ISS) and for the alcohol addicted patients with the Trier Alcoholism Scale (TIA). Both patient groups were matched with control groups containing 25 individuals with an analogous distribution in terms of age, sex and education. Results: As opposed to 44% of the alcohol addicted patients all PIU-patients (100%) fulfilled the criteria of another psychiatric disease, especially depression and anxiety disorders. As compared to their control groups, both the alcohol and the Internet addicted patients scored significantly higher in terms of depression (BDI), impulsivity (BIS) and inattention (CAARS). However, in none of the psychometric tests the two patient groups showed significant differences. Conclusions: Patients with PIU exhibit a clinical level of psychopathological symptomatology and share similar psychopathological and co-morbid features with alcohol addiction. In order to examine this novel form of addiction further and to treat patients adequately, it is argued that PIU should be established as a diagnostic entity in line with behavioral and substance addictions within diagnostic classification systems.

Keywords

Internet addiction; Alcoholism; Co-morbidity; Impulse control disorder; Pathological internet use

Introduction

With the digital turn, that has rapidly altered the globalized world like no media revolution before, the increasing occurrence of pathological Internet use (PIU), has become an issue of rising concern for more than a decade [1-3]. After initial skepticism within the medical community [4], now it can be hardly denied that the excessive or addicted use of the Internet, especially with online-gaming, has grown to a clinically relevant dimension of its own kind, challenging psychiatry’s diagnostic and therapeutic resources as predicted.

Internationally, the results of prevalence studies exhibit a wide rate due to methodological and cultural differences. In American studies prevalence of PIU has been found in 5.7% [1], in 8.1% [5] and in 0.3- 0.7% [6] of the examined populations. The tendency to lower but better i.e. more realistic estimates is reflected in recent studies from different countries exhibiting PIU-rates of 3.2% in German adults [7], 2,0% in Norwegian adolescents [8] and 1.6% in adolescents from South Korea [9]. In a British study with university students 18.3% were considered to be pathological Internet users [10]. In a similar study with US college students 4% scored in the occasionally problematic or addicted range on the Internet Addiction Test [11]. In a review Young and colleagues [12] estimate the general prevalence of adolescents between 4.6 and 4.7%, among college students 13-18.4% and in the general population 6 and 15%. Especially in the earlier studies, however, the lack of common criteria may have led to inaccurately high figures, raising the question, whether all of the subjects identified as pathological users actually suffered from a psychiatric condition of clinical relevance. Diagnostically, so far pathological Internet use (PIU) has been mostly viewed and examined with the presumption that it is a behavioral addiction [13] comparable not only to pathological gambling (PG) but also to alcohol addiction [14].

Within the context of DSM-IV [15] and ICD-10 [16] PIU still would have to be classified as an impulse control disorder (ICD). The diagnostic subgroup of ICD, however, is heterogeneous since it combines pathological ways of behavior such as pyromania and cleptomania with excessive non-pathological behavior, which are also labeled as behavioral addictions [17]. The most convincing diagnostic entity in this context is pathological gambling, to which the dependent use of the Internet can be referred best [18]. Accordingly, it would have to be labeled as pathological Internet use (PIU). Yet, from early on scientific approaches have claimed this phenomenon to represent an addiction [1,3,14]. In fact, pathological Internet use has already been under discussion to be listed and treated as an addiction in DSM-V as pathological gambling is already planned to be [19].

However, there remains a lack of consent about the question, if and how the addicted usage of the Internet is to be classified diagnostically. There are no studies yet directly comparing patients with PIU and patients with alcohol addiction. In this study, 25 PIU patients were compared clinically and psychometrically with 25 alcohol addicted patients to better understand the similarities and differences between these examples of substance and non-substance addiction. The study has stemmed from the hypothesis that PIU shares common psychopathological features with alcohol addiction, but has a higher rate of co-morbidity, which so far may have let it appear rather as a secondary phenomenon than as a diagnostic entity in itself. In this, the study may provide a contribution to the process of further determining the phenomenon of pathological Internet use diagnostically.

Materials and Methods

Participants

PIU subjects were responders to a public announcement of this study. In a preliminary telephone interview the study’s inclusion criteria were tested. Participants had to meet Young’s criteria for Internet addiction [20] as modified by Beard [21]. Subjects had to be self-motivated to seek therapeutic help. Only adult and physically healthy patients were accepted for enrollment. From the 32 eligible PIU-subjects 25 individuals completed the psychological testing. The study’s participants did not receive any financial incentive but could expect thorough diagnostic examination and both a concept and a referral for further psychiatric and psychotherapeutic treatment.

The alcohol dependent patients were recruited among freshly abstinent inpatients after the acute phase of withdrawal with a mean of 15 days of abstinence. From the 27 patients who met the inclusion criteria 25 subjects completed the study. In addition, two groups of 25 healthy control subjects were recruited for the patient groups with similar distribution of sex, age and school education and tested with the same psychometrical instruments.

Examinations

All examinations were done as face-to-face interviews. First, a freelance general psychiatric history and examination was taken with all alcohol and Internet dependent patients by the same specialist for psychiatry and psychotherapy. Second, the clinical diagnosis was confirmed or adjusted by the same scientist via Statistical Clinical Interview according to DSM-IV (SKID I), involving pre- and comorbidity [22]. Third, a test battery of psychometric instruments was performed as self reports to further investigate the patients’ media use and their psychopathological profile.

The German Internet Addiction Scale (ISS) was meant to measure the dimensions of pathological Internet use [7]. It is the best validated scale, at least in German-speaking countries for the examination of “Internet addiction”. The ISS consists of five subscales: Loss of control, withdrawal symptoms, tolerance, negative social consequences, consequences in terms of work and school performance. The ISS was composed from the translated item pool of the Internet Addiction Test, which was constructed by Young in English [3]. To better reflect the five subscales items were modified by Hahn and Jerusalem [7] with reference to the Internet User Survey [23] and the conceptualization by Orzack and Orzack [24]. The German Trier Alcoholism Inventory [25] confirmed the diagnosis of alcoholism and was correlated with the psychometric variables. The validated German version [26] of the Barrat Impulsiveness Scale (BIS) served to test the conceptualization of pathological Internet use as an impulse control disorder [27]. Derogatis’ Symptom Checklist (SCL-90R) was meant to screen for other psychopathological syndromes and to confirm the patients’ clinical level of distress [28,29]. The Beck Depression Inventory I (BDI) was used to test for depressive symptomatology [30,31]. And the German Version of the Conners Adult ADHD Rating Scales (CAARS) was used to identify the specific psychopathology of attention deficit and hyperactivity disorder [32,33].

Statistics

The case load estimate of 25 participants per group, calculated with nQuery Advisor 5.0 was based on the core psychometric instrument, the BIS. Data was assimilated and processed by the means of SPSS 16.0. T-tests for unrelated populations were performed between the two patient groups and their controls. Correlational calculations were done according to Pearson. A p-value ≤0.05 was considered to be significant.

Results

Socio-demographic variables and media use

The two examined patient groups were matched with two control groups in terms of sex, age and education (Table 1). Six (24%) of the internet dependent patients and their controls, and eight (32%) of the alcoholic patients and their controls were female. In terms of age and school education a match between the Internet- and alcoholdependent patients could not be achieved. Concerning age there was a significant difference (p≤0.001) between the PIU-patients (M=29.36; SD=10.76) and the alcoholics (M=41.36; SD=9.01). Descriptive statistics of the sample can be derived from Table 1.

Table 1: Socio-demographic data of the four study groups.

88% of controls in either of the two control groups and 60% of alcoholic patients used the internet at least regularly. The PIU patients used the Internet for non-work purposes significantly longer than the J Addict Res Ther Substance and Behavioral Addictions ISSN:2155-6105 JART, an open access journal control group and the alcohol addicted patients. As shown in Table 2 the use of games was also much higher in the PIU-group (61%) than in the group of alcoholics (28%), whose media use is also higher in general than their controls. However, these differences were not significant. The control groups’ general media use was less in all respects.

Table 2: Daily media use in PIU patients, alcoholics and controls.

Within the group of alcohol addicted patients Information seeking, E-mailing and chatting were the most frequent activities performed online, while in the PIU-group they were online-gaming (61%), especially role-playing (53% of online gamers).

Psychometric variables

In the Internet Addiction Scale (ISS) the PIU-group scored significantly (p≤0.001) higher (M=53.87; SD=13.35) than the alcohol group (M=24.83; SD=12.35) and the control-group I (M=23.79; SD=5.64). In the Barrat Impulsiveness Scale (BIS) both patient groups scored significantly higher than their control groups in the global scale and the subscale for inattention. However, in the other two BIS-subscales only the alcoholics showed significantly higher levels compared to their control group (Table 3).

Table 3: Internet addiction scores and impulsivity scores in PIU patients, alcoholics and controls.

In all global scores and subscales of the SCL-90R patients scored significantly higher than their respective controls, but there were no significant differences in any of the scales between the PIU and the alcoholic group (Table 4). For the PIU-patients the highest scores in the subscales were for depression (M=15,17; SD=6.06), insecurity (M=10.67; SD=6.12) and compulsivity (M=10.54; SD=6.27). For the alcoholic patients the ranking was quite similar: depression (M=16.13; SD=11.98), anxiety (M=11.50; SD=9.96) and compulsivity (M=10.58; SD=6.93). Within the PIU-group there have been no significant correlations between the ISS and the SCL-90R subscales. Within the alcohol group the TAI score, representing the severity of alcoholism, correlated positively with the subscales for depression (p≤0.05; r=.411) and insecurity (p≤0.05; r=.478).

Table 4: Symptom-Checklist (SCL-90R)results in PIU patients, alcoholics and controls.

In all scales of the CAARS for ADHD both patient groups showed significantly higher scores than their control groups, except for the CAARS-subscale for hyperactivity in PIU-patients (Table 5). Compared to the alcoholics, the PIU-group exhibited slightly but not significantly higher results in all CAARS-scales except for the one for hyperactivity.

Table 5: Results of ADHS and depression screening in PIU patients, alcoholics and controls.

The Beck Depression Inventory (BDI) identified significantly more depressive psychopathology in both the PIU and the alcohol patients as compared to their control groups (p≤0.001), yet without significant differences between patient groups. Judging by the BDI’s cut off score 80% of the PIU-patients and 64% of the alcoholics were moderately or severely depressed (Table 6).

Table 6: Beck Depression Inventory (BDI) classificationin internet-dependent patients, alcoholics and controls.

Correlations between the psychometric instruments and ISS/ TAI are shown in Table 7. The ISS significantly correlated with the CAARS global scale (p≤0.05; r=.462) and its subscales for hyperactivity, (p≤0.05; r=.513) and impulsivity (p≤0.05; r=.464). The TAI significantly correlated with the BDI (p≤0.05; r=.441) and with the three SCL-90R global scores for Severity (p≤0.01; r=.538), Number of Symptoms (p≤0.005; r=.558) and Intensity of Response (p≤0.05; r=.488).

Table 7: Correlations of the internet addiction scale ISS (for PIU-patients) and TAI (for alcoholics) with the main psychometric instruments.

Clinical results

Twelve (48%) of the 25 PIU-patients and eight (32%) of the 25 alcohol dependent patients reported to have suffered from depression beforehand. Six PIU patients (24%) and four alcoholics (16%) have previously been diagnosed with an anxiety disorder. Six patients with PIU (24%) and four alcoholics (16%) reported to have tried to commit suicide. In Table 8 the co-morbid psychiatric diseases are presented. Employing the SCID all PIU-patients were diagnosed with another psychiatric disorder, while this was the case in only 44% of the alcohol dependent patients. In both groups, disorders with a depressive syndrome were the most frequent diagnostic entities. The vast majority of the alcoholics (92%) and twelve of the 25 PIU-patients (48%) were smokers. Two of the Internet addicted patients have been formerly abused alcohol, one cannabis and one cocaine. At the time of examination, none of the PIU-patients was abusing illicit drugs.

Table 8: Co-morbid psychiatric diagnoses in PIU and alcohol-dependent patients.

Discussion

Socio-demographic variables and media use

Because of limited sample size and the selective inclusion of participants conclusions from the socio-demographic variables must be drawn with caution. The fact that the level of education of the Internet addicted individuals has been comparably high was puzzling. Several other studies with children and adolescents in societies with high availability of information infrastructure have instead shown a link between poor performance in school and Internet- and videogame addiction, yet results differ concerning the question, whether PIU it is rather a cause [34] or an effect [35] of school education failure. However, among these it can be expected that those with higher education had a better access to social support.

Less than half of the PIU-patients worked as compared to 80% of their matched control group. This result is in line with those of Griffiths and Wood [35] who documented the link between excessive media use and problems or failure in professional training.

Except for sex distribution, alcohol dependent patients were matched poorly to the PIU-patients, being significantly older and less educated. Only 36% of the alcohol dependent patients were employed. It can be derived that all patients suffered a decisive impact on their professional career. The data may suggest, that this effect is even more dramatic in PIU patients since they are equally affected by the psychosocial impact of the disease at an earlier age and in spite of higher educational level.

The PIU-patients used the Internet significantly longer hours than the alcoholics (p≤0.01). PIU-patients were mostly occupied with online-gaming, especially with Massively Multiplayer Online Role- Playing Games (MMORPGs), which has been shown by similar studies as well [18,36]. By contrast, the control groups and the alcohol addicted patients, whose media use was only slightly higher, were using the Internet mostly for Information seeking, E-mailing and Chatting. The specific addictive effects of Internet-games, especially of MMORPGs have been detected only recently. These games share addictive features known from gambling [37], especially due to their intermittent reward systems.

Psychometric variables and clinical results

The Internet Addiction Scale (ISS) confirmed highly significant differences between the PIU-group and both its control group and the alcohol group indicating that the examined individuals really suffered from Internet addiction. Since the ISS-subscales identified typical addictive features in the PIU-patients such as loss of control, withdrawal, tolerance, social and professional consequences, it may be argued that PIU resembles disorders of substance abuse and addiction [38].

To test the classification of PIU as an impulse control disorder the Barrat Impulsiveness Scale (BIS) was used, showing very similar results in PIU patients and alcohol dependent patients, with significant differences to their control groups. The BIS identified 68% of the PIU-patients and 56% of the alcoholic patients as highly impulsive. This result may be seen as another hint that impulsivity may not only be a characteristic symptom of PIU but of addiction in general [39], although the BIS and its subscales did not correlate with the instruments measuring the intensity of the specific dependency within the two patient groups. Yet, it must be underlined that the difference in impulsivity between the PIU-patients and their control group mainly derives from a higher value in the BIS-subscale for inattention in the PIU-group. Conclusively, this data may contribute to the notion that PIU and other behavioral addictions may be not sufficiently characterized by the nosological category impulse control disorder, which has been put into critical focus altogether [40].

The results of the Conners Adult ADHD Rating Scales (CAARS) imply similar interpretations. The Internet and alcohol dependent patients did not reveal any significant differences but scored significantly more pathological than their control groups, both globally and in nearly all subscales, highly significant in the scales for inattention and self-control. Moreover, within the PIU-group the global CAARS ADHD Index and the subscores for hyperactivity and impulsivity correlate significantly with the ISS. This may contribute to the hypothesis that not only for substance abuse [41,42] but also for non-substance addiction like pathological media use persisting ADHD in adulthood may be a risk factor. However, the influence of excessive media use has been an issue of controversial discussion within the field of ADHD research [43] and although there is some evidence for causal relationships in both ways [44,45], no proven link can be claimed yet [46].

Freshly gained abstinence in the group of alcohol-dependent patients may have had a confounding effect limiting the comparability of the two groups with regard to SCL-90 subscales, especially because they are referring to the psychopathological symptoms in the last seven days. As can be seen from the results, there were striking differences between patient and control groups in many psychometric measures but rarely between each other. The alcoholic patients seemed to suffer more from anxiety and phobia whereas the PIU patients showed higher results in insecurity, which may be an effect of the age differences between the groups. In a Korean study by Ha and colleagues [47] it has been shown that the co-morbid psychiatric disorders occurring with PIU are indeed age-related. And the highest scores in both patients groups and the most significant difference to the control groups were in the depression subscale.

These results are also reflected in the highly-significant pathological BDI scores of the two patient groups. The BDI, however, correlated significantly only among alcohol addicted patients with the severity (TAI) of alcohol addiction [48] but not among PIU patients with the respective severity (ISS). Although it has to be said that depressiveness is a syndrome, which goes along with many psychiatric diseases, the data suggests that it also plays an important role in nonsubstance abuse disorders such as PIU.

The main clinical result of the study was that all Internet dependent patients (100%) as opposed to only 44% of the alcohol dependent patients were diagnosed with another psychiatric disorder, with 76% of the PIU patients suffering from major depression (36% of the alcohol dependent patients). Anxiety disorders, sometimes occurring as a third psychiatric diagnosis, occur somewhat less often, in both patient groups. In a similar German study by Kratzer and Hegerl [49] 90% of 30 PIU patients did exhibit another psychiatric disease as diagnosed with the Munich Composite International Diagnostic Interview (CIDI), but showing more anxiety than depressive pathology.

General discussion

The presented data may confirm the results of similar studies that psychiatric co-morbidity is a common feature of pathological Internet use [47,49]. The fact that all patients were diagnosed with another psychiatric disease may partly be due to the profound diagnostics, which partly constitute an examiner bias. Also, we may only have seen the tip of the iceberg. Study PIU-patients were severely affected with an average private Internet use of almost 7 h/d. Apart from that, the frequent co-morbidity of PIU with depression and anxiety disorders resembles the typical co-morbid spectrum of substance abuse disorders, especially of alcohol dependency [50,51]. Also the psychometric similarities between the Internet and alcohol dependent patients in impulsiveness, inattention and depression may contribute to the notion that PIU shares common features with substance abuse disorders. It may be important to establish Internet addiction as a distinct diagnostic entity to provide adequate treatment of these patients and to benefit from the longtime knowledge and experience of addiction research and therapy.

There are several weaknesses in the study’s design, which limit the interpretability of its results. The comparability of the two patient populations was limited since they differed significantly in age and school education. This difference led to the necessity that two separate control groups had to be recruited. Also, the number of 25 individuals per group is low, especially when more information should be derived about co-morbid psychopathology.

More research with larger populations is needed to better understand the nature of PIU, its risk factors, underlying psychopathology and cause-effect relations. High rates of co-morbidity make it obligatory to address the question, whether PIU may be seen as a sole symptom of a primary disorder. This may be the case to a varying degree. However here are other substance abuse disorders with a rate of comorbidity nearing 100% (like heroin addiction), which does not imply that they are just a symptom. While excessive consumption of internet pornography and pathological gambling did not confound our sample, they can cause states of excessive internet use not to be classified as PIU. More emphasis should be put on the question, which media forms and contents have stronger addictive potential, in order to derive preventive measures soon, taking into account that media addiction often already stems from childhood and adolescence.

Author Disclosure Statement

None of the contributing authors declare any competing financial interests. The study has been financed fully by the Hanover Medical School’s intramural research fund. There have been no commercial products involved. The authors have not accepted funding from the gambling, gaming, tobacco or alcohol industry..

References

1. Greenfield DN (1999)
2. Griffiths MD, Hunt N (1998)
3. Young KS (1996)
4. Mitchell P (2000)
5. Morahan-Martin J, Schumacher P (2000)
6. Aboujaoude E, Koran LM, Gamel N, Large MD, Serpe RT (2006)
7. Hahn A, Jerusalem M (2001) Internetsucht: Jugendliche gefangen im Netz. In: Raithel J, editor. Risikoverhaltensweisen Jugendlicher Erklärungen, Formen und Prävention.
8. Johansson A, Gotestam KG (2004)
9. Kim K, Ryu E, Chon MY, Yeun EJ, Choi SY, et al. (2006)
10. Niemz K, Griffiths M, Banyard P (2005)
11. Christakis DA, Moreno MM, Jelenchick L, Myaing MT, Zhou C (2011)
12. Young KS, Aue XD, Ying L (2010)
13. Herpertz S, Sass H (1997)
14. Bai YM, Lin CC, Chen JY (2001)
15. American Psychiatric Association (1994)
16. WHO (1992)
17. Albrecht U, Kirschner NE, Grusser SM (2007)
18. Wolfling K, Buhler M, Lemenager T, Morsen C, Mann K (2009)
19. Miller G, Holden C (2010)
20. Young KS (1999) Internet Addiction: Symptoms, Evaluation, And Treatment. In: Van de Creek L, Jackson X, editors. Innovations in Clinical Practice: a source book Sarasota: Professional Resource Press 19-31.
21. Beard KW, Wolf EM (2001)
22. Steinberg M, Buchanan DJ, Cicchetti D, Hall P (1993) Clinical assessment of dissociative symptoms and disorders: The Structured Clinical Interview for DSM-IV Dissociative Disorders (SCID-D). Dissociation 6: 3-15.
23. Brenner V (1997)
24. Orzack MH, Orzack DS (1999)
25. Funke W, Funke J, Klein M, Scheller R (1987) Trierer Alkoholismus-Inventar. Hogrefe, Göttingen, Germany.
26. Preuss UW, Rujescu D, Giegling I, Koller G, Bottlender M, et al. (2003)
27. Patton JH, Stanford MS, Barratt ES (1995)
29. Derogatis LR, Cleary PA (1977)
30. Hautzinger M (1991)
31. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J (1961)
32. Rosler M, Retz W, Retz-Junginger P, Stieglitz RD, Kessler H, et al. (2008)
33. Erhardt D, Epstein JN, Conners CK, Parker JDA, Sitarenios G (1999)
34. Yang SC, Tung CJ (2004)
35. Griffiths M, Wood RT (2000)
36. Smahel D, Blinka L, Ledabyl O (2008)
37. Ng BD, Wiemer-Hastings P (2005)
38. Hasin D, Hatzenbuehler ML, Keyes K, Ogburn E (2006)
39. Westphal J (2007)
40. Dell'Osso B, Altamura AC, Allen A, Marazziti D, Hollander E (2006)
41. Ohlmeier MD, Peters K, Kordon A, Seifert J, Wildt BT, et al. (2007)
42. Ohlmeier MD, Peters K, Te Wildt BT, Zedler M, Ziegenbein M, et al. (2008)
43. Chan PA, Rabinowitz T (2006)
44. Yen JY, Ko CH, Yen CF, Wu HY, Yang MJ (2007)
45. Stevens T, Mulsow M (2006)
46. Christakis DA, Zimmerman FJ, DiGiuseppe DL, McCarty CA (2004)
47. Ha JH, Yoo HJ, Cho IH, Chin B, Shin D, et al. (2006)
48. Korkeila J, Kaarlas S, Jääskeläinen M, Vahlberg T, Taiminen T (2010)
49. Kratzer S, Hegerl U (2008)
50. Arolt V, Driessen M (1996)
51. Yen JY, Ko CH, Yen CF, Chen SH, Chung WL, et al. (2008)