Journal of Affective Disorders 57 (2000) 267–272 www.elsevier.com / locate / jad

Preliminary communication

Psychiatric features of individuals with problematic internet use a, a b Nathan A. Shapira M.D., Ph.D. *, Toby D. Goldsmith M.D. , Paul E. Keck Jr. M.D. , b b Uday M. Khosla , Susan L. McElroy M.D. a

b

Department of Psychiatry, University of Florida, P.O. Box 100256, Gainsville, FL 32610 -0256, USA Department of Psychiatry, University of Cincinnati Medical Center, P.O. Box 670559, Cincinnati, OH 45267 -0559, USA Received 15 February 1999; accepted 1 June 1999

Abstract Background: Problematic internet use has been described in the psychological literature as ‘internet addiction’ and ‘pathological internet use’. However, there are no studies using face-to-face standardized psychiatric evaluations to identify behavioral characteristics, psychiatric comorbidity or family psychiatric history of individuals with this behavior. Methods: Twenty individuals with problematic internet use were evaluated. Problematic internet use was defined as (1) uncontrollable, (2) markedly distressing, time-consuming or resulting in social, occupational or financial difficulties and (3) not solely present during hypomanic or manic symptoms. Evaluations included a semistructured interview about subjects’ internet use, the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders-IV (SCID-IV), family psychiatric history and the Yale–Brown Obsessive–Compulsive Scale (Y-BOCS) modified for internet use. Results: All (100%) subjects’ problematic internet use met DSM-IV criteria for an impulse control disorder (ICD) not otherwise specified (NOS). All 20 subjects had at least one lifetime DSM-IV Axis I diagnosis in addition to their problematic internet use (mean6SD 5 5.163.5 diagnoses); 14 (70.0%) had a lifetime diagnosis of bipolar disorder (with 12 having bipolar I disorder). Limitations: Methodological limitations of this study included its small sample size, evaluation of psychiatric diagnoses by unblinded investigators, and lack of a control group. Conclusions: Problematic internet use may be associated with subjective distress, functional impairment and Axis I psychiatric disorders.  2000 Elsevier Science B.V. All rights reserved. Keywords: Internet; OCD; Impulse control disorder; Bipolar

1. Introduction Problematic internet use, wherein an individual’s inability to control his or her use of the internet *Corresponding author. Tel.: 1 1-352-392-2831; fax: 1 1-352392-2579.

causes marked distress and / or functional impairment, has been described in the psychological literature as ‘internet addiction’ and ‘pathological internet use’, based on the DSM-IV definition for substance dependence and pathological gambling, respectively (Griffiths, 1996, 1997; O’Reilly, 1996; Stein, 1997; Young, 1996, 1998; Young and Rogers, 1998).

0165-0327 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0165-0327( 99 )00107-X

268

N. A. Shapira et al. / Journal of Affective Disorders 57 (2000) 267 – 272

Proposed treatments for afflicted individuals have included time-management techniques (Young, 1998) and participation in support groups (O’Reilly, 1996; Young, 1998). Persons with problematic internet use have been reported to have high rates of depressive symptoms. An on-line survey utilizing the World Wide Web (WWW) and the Beck Depression Inventory in 259 individuals classified as having pathologic internet use demonstrated mild to moderate depression with a mean (SD) of 11.2 (13.9) (Young and Rogers, 1998). In addition, a recent prospective study of 169 subjects demonstrated an association between excessive internet use and depression (even when previous depression scores were controlled for), loneliness, declines in communication with household family members and social withdrawal (Kraut et al., 1998). However, the relationship between problematic internet use and formal mental disorders is unknown as there have been no studies utilizing standardized structured interviews to identify behavioral characteristics, co-occurring Axis I disorders, or family history of psychiatric illness in individuals displaying this behavior (Stein, 1997). Over the last 15 years, a broad range of pathological repetitive behaviors have been hypothesized as possibly being related to obsessive–compulsive disorder (OCD), and are referred to as obsessive– compulsive (or OCD) spectrum disorders (Goldsmith et al., 1998). Many of these behaviors are associated with comorbid OCD and depressive disorders and possible preferential response to serotonin re-uptake inhibitors (SRIs). Examples include body dysmorphic disorder (BDD), hypochondriasis, anorexia nervosa, Tourette’s syndrome and possibly some impulse control disorders (ICDs). We hypothesized that problematic internet use might be another form of an OCD spectrum disorder and, in addition to repetitive behavior, be characterized by obsessive thinking, comorbidity with OCD and depressive disorders, and preferential response to SRIs. To further study individuals with problematic internet use, we systematically evaluated a consecutive group of users whose internet usage met criteria as described below utilizing face-to-face structured diagnostic interviews and rating instruments. We also systematically assessed treatment response histories in those subjects receiving prior treatment.

2. Methods

2.1. Subjects Men and women, 18 years of age or older, who had problematic internet use for at least six months were recruited. Problematic internet use was defined as (a) uncontrollable, (b) markedly distressing, timeconsuming or resulting in social, occupational or financial difficulties and (c) not solely present during hypomanic or manic symptoms. Subjects provided written informed consent to be interviewed.

2.2. Assessments Evaluations included a semistructured interview to assess demographic information, the nature of internet use, including time spent on ‘essential’ (required job / school functions) versus ‘nonessential’ (pleasure / recreational or personal) use, and a retrospective rating of response of problematic internet use to any previous mental health treatment. Response of problematic internet use was defined as: none (0– , 25% decrease in internet usage or urges to use), mild (25– , 50% decrease), moderate (50– , 75% decrease) or marked (75–100% decrease). Evaluations also included the Structured Clinical Interview for DSM-IV–Patient Version (SCID-IV) augmented with additional DSM-IV modules for ICDs not otherwise specified (NOS), paraphilias and somatoform disorders (available from the authors upon request) to assess Axis I psychiatric comorbidity, family history evaluation for psychiatric disorders in first- and second-degree relatives done via the family history method (Andreasen et al., 1977) and the Y-BOCS (Goodman et al., 1989) modified for internet use to assess the obsessive–compulsive nature of the use.

3. Results The study sample consisted of 20 subjects (eleven men and nine women) with a mean6SD of 36.0612.0 years of age. Twelve subjects (60.0%) responded to newspaper advertisements for problematic internet use and eight (40.0%) were clinically referred with a ‘chief complaint’ of problematic

N. A. Shapira et al. / Journal of Affective Disorders 57 (2000) 267 – 272

internet use. Age of onset of problematic use was 33.3612.2 years. All subjects completed high school or had a GED equivalent and 17 (85.0%) had attended college. Fourteen (70.0%) subjects were employed. Thirteen (65.0%) had used computers prior to the availability of the internet, with two individuals reporting problematic computer use prior to availability of the internet. ‘Essential’ use was (mean6SD) 2.864.8 hours per week, but ‘nonessential’ use was 27.9620.5 hours per week. ‘Nonessential’ use occurred in the following domains: chat forums (17.7%), e-mail (15.1%), WWW ‘surfing’ (14.0%), multi-user domains (12.7%), miscellaneous uses including games and designing web pages (9.2%), pornography (8.1%), news / current events (6.8%), newsgroups (6.1%), file transfers (3.7%), music (3.2%), shopping / buying (2.5%), card catalogs (0.6%) and political uses (0.3%). Problems associated with internet use were: significant social impairment (e.g., family strife or divorce) in 19 (95.0%) subjects, marked personal distress over their behaviors in 12 (60.0%) subjects, vocational impairment (e.g., failing in college, decreased job productivity or loss of job) in eight (40.0%) subjects, financial impairment (e.g., substantial debt due to heavy use of the internet) in eight (40.0%) subjects and legal problems (e.g., caught harassing someone online) in two (10.0%) subjects. An unexpected finding was that every subject’s problematic internet use (100%) met DSM-IV criteria for an ICD NOS. By contrast, only three subject’s (15.0%) problematic internet use meet DSM-IV criteria for OCD. The average (mean6SD) Y-BOCS score modified for internet use was 19.665.6. Consistent with subjects meeting the DSM-IV definition of an ICD (as opposed to OCD), the Y-BOCS question concerning distress associated with thoughts, impulses or urges to use the internet was rated lowest of all of the ten questions (mean6SD 5 0.660.7), indicating a low level of distress. Furthermore, subjects rated the Y-BOCS question concerning degree of effort made in resisting the excessive internet use the highest of all ten questions (mean6SD 5 2.96.9), indicating a low level of resistance. All subjects met criteria for at least one lifetime DSM-IV Axis I diagnosis (mean6SD 5 5.163.5 diagnoses) (Table 1). As expected, mood and anxiety

269

disorders were the most common associated Axis I disorders. However, to our surprise, bipolar disorders (as opposed to depressive disorders) were the most common mood disorders, and phobias (as opposed to OCD) were the most common anxiety disorders. Nineteen (95.0%) of 20 individuals had histories of psychiatric disorders in family members. Thirteen (65.0%) had at least one first- or second-degree relative with a depressive disorder, ten (50.0%) with a bipolar disorder and 12 (60%) with a substance use disorder. Seventeen (85.0%) subjects had received previous mental health treatment and 15 (75.0%) had received treatment with psychotropic medications. As shown in Table 2, 5 (35.7%) of 14 antidepressant monotherapy trials resulted in moderate or marked reduction in problematic internet use, with only two (22.2%) of nine selective serotonin re-uptake inhibitor (fluoxetine, paroxetine and sertraline) trials resulting in favorable responses. By contrast, 14 (58.3%) of 24 single- or combination-agent mood stabilizer trials were associated with favorable responses. Moreover, this favorable response rate increased to 12 (75.0%) of 16 trials when trials involving concurrently administered antidepressants or stimulants were excluded from analysis.

4. Discussion In this small group of subjects, problematic internet use was found to be associated with subjective distress, considerable social, vocational and / or financial impairment, as well as substantial psychiatric comorbidity. These findings are consistent with other studies reporting depressive symptoms and social impairment in excessive internet users (Kraut et al., 1998; Young and Rogers, 1998). However, we had several unexpected findings. First, the problematic internet use of this small group was more impulsive and ego syntonic than compulsive and ego dystonic in nature, and more closely resembled the DSM-IV definition of an ICD than that of OCD. Individuals described an increasing sense of tension or arousal before successfully logging onto the internet, which was difficult or impossible to resist, and a relief of that tension that was often pleasurable as they logged on. Of note, this observa-

270

N. A. Shapira et al. / Journal of Affective Disorders 57 (2000) 267 – 272

Table 1 DSM-IV Axis I diagnoses in 20 subjects with problematic use of the internet Diagnosis

Current diagnosis

Lifetime diagnosis

n

%

n

%

14 2 11 1

70.0 10.0 55.0 5.0

17 3 12 2

85.0 15.0 60.0 10.0

Psychotic disorders Schizoaffective disorder, bipolar type

2 2

10.0 10.0

2 2

10.0 10.0

Substance use disorders a Alcohol abuse or dependence Other substance abuse or dependence

2 2 1

10.0 10.0 5.0

11 9 9

55.0 45.0 45.0

12 2 0 4 8 3 5 2

60.0 10.0 0.0 20.0 40.0 15.0 25.0 10.0

14 5 1 7 9 4 5 2

70.0 25.0 5.0 35.0 45.0 20.0 25.0 10.0

Paraphilias a Exhibitionism Sexual Masochism Sexual Sadism Transvestic fetishism Voyeurism

3 1 1 2 1 1

15.0 5.0 5.0 10.0 5.0 5.0

3 1 1 2 1 1

15.0 5.0 5.0 10.0 5.0 5.0

Eating disorders Anorexia nervosa Bulimia nervosa Eating disorder NOS b

3 0 1 2

15.0 0.0 5.0 10.0

7 1 2 4

35.0 5.0 10.0 20.0

Impulse control disorders a Intermittent explosive disorder Kleptomania Pathological gambling Impulse-control disorder NOS c

7 2 1 1 4

35.0 10.0 5.0 5.0 20.0

10 3 2 1 6

50.0 15.0 10.0 5.0 30.0

Other Body dysmorphic disorder

2

10.0

2

10.0

Mood disorders Major depression Bipolar I Bipolar II

Anxiety disorders a Panic disorder with or without agoraphobia Agoraphobia Specific phobia Social phobia Obsessive–compulsive disorder Post-traumatic stress disorder Generalized anxiety disorder

a

Total is less than the sum of disorders; patients had more than one disorder in category. All patients had binge eating disorder. c All patients had compulsive buying. b

tion is similar to the report of an individual with ‘compulsive computer use’ who also had excessive use of the internet (as well as video games when he did not have access to the internet) (Belsare et al., 1997). Second, nearly all individuals (80.0%) met lifetime criteria for bipolar disorder or schizoaffective disorder, bipolar type, and there were comparatively low lifetime rates of comorbid major depressive disorder (15%) and OCD (20%). Third, in-

dividuals with problematic internet use appeared to show a preferential treatment response to mood stabilizers over antidepressants. Taken together, these findings were contrary to our original hypothesis that these individuals would demonstrate high degrees of compulsivity, high rates of OCD and depressive disorders, and a favorable response to SRI antidepressants. However, our finding that problematic internet use resembles an ICD and is associ-

N. A. Shapira et al. / Journal of Affective Disorders 57 (2000) 267 – 272

271

Table 2 Responses of 15 subjects with problematic use of the internet to psychotropic medications Medication Antidepressant monotherapy Fluoxetine Paroxetine Mirtazapine Sertraline Nefazodone Venlafaxine Bupropion Mood stabilizer monotherapy Lithium Gabapentin Divalproex Mood stabilizer dual-therapy Antipsychotic monotherapy Anxiolytic monotherapy Antidepressant / anxiolytic combination Mood stabilizer / antidepressant combination Mood stabilizer / antipsychotic combination Mood stabilizer / anxiolytic combination Mood stabilizer / stimulant combination Other therapy b Other therapy c

Number of trials

Favorable response a (no. of patients)

3 2 1 4 1 2 1

0 2 1 0 0 1 1

2 1 4 1 1 3 1 5 7 1

2 1 1 1 1 1 0 1 6 1

1

0

1 1

1 0

a

Favorable response defined as moderate or marked response in terms of decreased internet usage. Other therapy 5 divalproex, buspirone and imipramine combination. c Other therapy 5 lithium and hypericum combination. b

ated with bipolarity is consistent with earlier observations that ICDs and bipolar disorder may be related due to similarities in phenomenology, comorbidity, family psychiatric history and treatment response (McElroy et al., 1996). These findings must be viewed critically, as this is a preliminary study with many limitations. These include its small size, self-reported interviews, unblinded investigators, lack of a control group, and the possibility of overestimating certain psychiatric disorders, particularly bipolar disorder. However, several aspects of the data support the findings of the high rate of bipolarity. First, although 60.0% of subjects were self-referred for evaluation of their internet use, many had already received psychiatric diagnoses and psychiatric care, specifically for bipolar disorder. For example, of the 12 individuals who received the SCID diagnosis of lifetime bipolar I disorder, eight (66.7%) had received previous treatment with mood stabilizers and / or antipsychotics. Moreover, one of

the other four individuals subsequent to her evaluation required treatment with a mood stabilizer after experiencing a mixed manic episode while receiving an antidepressant. Second, individuals displayed a high rate of psychiatric disorders in their families (95%), with ten (50%) individuals having at least one relative with bipolar disorder. Five of these ten subjects had multiple relatives with bipolar disorder. Furthermore, the family history method has been shown to underestimate the degree of psychiatric illness in family members (Andreasen et al., 1977). Third, more individuals retrospectively reported moderate or marked improvement in control over their internet use that was better with mood stabilizers than with antidepressants. It is important to note, however, that of the 12 individuals with a current diagnosis of bipolar I or II disorder, all had or most recently experienced a depressed or mixed episode; none met criteria for a current or recent manic episode. Moreover, when

272

N. A. Shapira et al. / Journal of Affective Disorders 57 (2000) 267 – 272

evaluating subjects, they would spontaneously discuss their depressed symptoms, whereas their hypomanic and manic symptoms usually became apparent only with standardized structured interviews. This might explain why previous researchers utilizing online surveys, which focus on depressive but not manic symptoms, have failed to report bipolarity in persons with problematic internet use. Indeed, over the last several years, there has been greater appreciation that the true bipolar phenotype may be much broader than how it is defined in DSM-IV and, hence, that bipolar disorder in all its many forms may be far more common than realized in populations with depressive symptoms (Akiskal, 1996). For example, when hypomania was systematically searched for in patients with at least one DSM-IV major depressive episode utilizing a French multicenter study (EPIDEP), 40% of patients with a major depressive episode(s) were classified as having bipolar II disorder (Hantouche et al., 1998). Only half of these patients were known bipolar II patients to the clinicians at study entry (Hantouche et al., 1998). Another example includes the evaluation of 108 consecutive anxious and / or depressed patients in a naturalistic family practice setting, in which 28 (25.9%) were diagnosed with a bipolar spectrum disorder (e.g., bipolar I, II or III disorder or cyclothymia) (Manning et al., 1997). In summary, problematic internet use may represent a clinically important syndrome that is associated with distress, functional impairment and psychiatric disorders. A more precise understanding of the psychopathology and psychiatric comorbidity of problematic internet use would help delineate whether it is a distinct disorder (e.g., an ICD), a symptom of an already characterized psychiatric illness (e.g., a mood disorder, especially bipolar disorder), or both.

References Akiskal, H.S., 1996. The prevalent clinical spectrum of bipolar disorders: beyond DSM-IV. J. Clin. Psychopharmacol. 16 (2 Suppl 1), 4S–14S.

Andreasen, N.C., Endicott, J., Spitzer, R.L., Winokur, G., 1977. The family history method using diagnostic criteria: inter-rater reliability and validity. Arch. Gen. Psychiatry 34, 1229–1235. Belsare, T.J., Gaffney, G.R., Black, D.W., 1997. Compulsive computer use (letter). Am. J. Psychiatry 154, 289. Griffiths, M.D., 1996. Internet addiction: an issue for clinical psychology? Clin. Psychology Forum 97, 32–36. Griffiths, M.D., 1997. Psychology of computer use: XLIII. Some comments on ‘addictive use of the internet’ by Young (comment). Psychological Rep. 80, 81–82. Goldsmith, T.D., Shapira, N.A., Phillips, K.A., McElroy, S.L., 1998. Obsessive compulsive spectrum disorders. In: Swinson, R.P., Antony, M.M., Rachman, S., Richter, M.A. (Eds.), Obsessive–Compulsive Disorder: Theory, Research, and Treatment, Guilford Publications, New York, pp. 397–425. Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C., Fleischmann, R.L., Hill, C.L., Heninger, G.R., Charney, D.S., 1989. The Yale–Brown obsessive compulsive scale. I. Development, use, and reliability. Arch. Gen. Psychiatry 46, 1006–1011. Hantouche, E.G., Akiskal, H.S., Lancrenon, S., Allilaire, J.F., Sechter, D., Azorin, J.M., Bourgeois, M., Fraud, J.P., ChatenetDuchene, L., 1998. Systematic clinical methodology for validating bipolar-II disorder: data in mid-stream from a French national multi-site study (EPIDEP). J. Affect. Disord. 50, 163–173. Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukopadhyay, T., Scherlis, W., 1998. Internet paradox: a social technology that reduces social involvement and psychological well-being? Am. Psychol. 53, 1017–1031. Manning, J.S., Haykal, R.F., Connor, P.D., Akiskal, H.S., 1997. On the nature of depressive and anxious states in a family practice setting: the high prevalence of bipolar II and related disorders in a cohort followed longitudinally. Compr. Psychiatry 38, 102–108. McElroy, S.L., Pope, H.G., Keck, Jr. P.E., Hudson, J.I., Phillips, K.A., Strakowski, S.M., 1996. Are impulse-control disorders related to bipolar disorder? Compr. Psychiatry 37, 229–240. O’Reilly, M., 1996. Internet addiction: a new disorder enters the medical lexicon. Can. Med. Assoc. J. 154, 1882–1883. Stein, D.J., 1997. Internet addiction, internet psychotherapy [letter; comment]. Am. J. Psychiatry 153, 890. Young, K.S., 1996. Psychology of computer use: XL. Addictive use of the internet: a case that breaks the stereotype. Psychological Rep. 79, 899–902. Young, K.S., 1998. Caught in the Net: How to Recognize the Signs of Internet Addiction — and A Winning Strategy For Recovery, Wiley, New York. Young, K.S., Rogers, R.C., 1998. The relationship between depression and internet addiction. CyberPsychology and Behavior 1, 25–28.