Untitled

Crystal Methamphetamine Use Predicts Incident STD InfectionAmong Men Who Have Sex With Men Recruited Online: A NestedCase-Control Study Sabina Hirshfield1, PhD; Robert H Remien2, PhD; Imelda Walavalkar1, BA; Mary Ann Chiasson1, DrPH 1Medical and Health Research Association of New York City, Inc, New York, NY, USA2HIV Center for Clinical and Behavioral Studies, New York State Psychiatric Institute and Columbia University, New York, NY, USA Corresponding Author:
Sabina Hirshfield, PhD
Medical and Health Research Association of New York City, Inc.
40 Worth Street, Suite 720
New York, NY 10013
USA
Phone: +1 212 285 0220 ext 115
Fax: +1 212 385 0565
Email:
Background: Among men who have sex with men (MSM), the number of newly diagnosed human immunodeficiency virus
(HIV) infections has increased by approximately 60% since 1999. Factors that may be contributing to this resurgence include a
widely reported increase in bacterial sexually transmitted diseases (STDs) among HIV-positive and HIV-negative MSM, as well
as unsafe sexual practices.
Objective: This research was undertaken to learn more about risk behaviors associated with an incident STD among MSM.
Methods: A nested case-control study was conducted, using data from a cross-sectional Internet survey of MSM (N=2643),
which investigated risk behaviors during a 6-month period in 2001. Chi-square and logistic regression methods were used to
estimate the likelihood of acquiring an incident STD versus no STD.
Results: Eighty-five percent of the respondents were white, 46% were under age 30, and 80% had met sex partners online; 7%
were HIV-positive. Men with an incident STD were more likely than men without an STD to report drug use (crystal
methamphetamine odds ratio 3.8; 95% confidence interval 2.1-6.7; cocaine OR 2.3; 95% CI 1.2-4.2; ecstasy OR 2.2; 95% CI
1.3-3.8; Viagra OR 2.1; 95% CI 1.2-3.7), alcohol before or during sex (OR 1.9; 95% CI 1.2-2.9), and high-risk sexual behavior
(unprotected anal intercourse OR 5.0; 95% CI 2.8-8.9; multiple sex partners OR 5.9; 95% CI 2.5-13.8). In the multivariate analysis,
significant independent predictors associated with an incident STD were crystal methamphetamine use (adjusted OR 2.0; 95%
CI 1.1-3.8), unprotected anal intercourse (adjusted OR 3.4; 95% CI 1.9-6.3), and 6 or more sex partners during the study period
(adjusted OR 3.3; 95% CI 1.4-7.8).
Conclusion: Identifying and treating MSM who have STDs, or who are at increased risk for acquiring STDs, is crucial in
preventing the further spread of disease. In addition, there is a need to integrate HIV/STD and substance use prevention and
education into Web-based and community-based venues.
(J Med Internet Res 2004;6(4):e41)
KEYWORDS:
Internet; sexually transmitted diseases; methamphetamine; HIV sexually transmitted diseases (STDs), namely syphilis andgonorrhea, among MSM [,]. Not only are STDs a marker for Among men who have sex with men (MSM), the number of unsafe sexual behavior, but ulcerative and non-ulcerative STDs newly diagnosed human immunodeficiency virus (HIV) increase HIV viral load and infectivity in persons with HIV Several factors may be contributing to the increase in HIV []. Moreover, studies have found a high proportion of transmission. One is the widely reported increase in bacterial J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.1 continued unsafe sexual practices and exposure of others to between June and December 2001, as part of a larger study of HIV. Substance use has also been associated with sexual risk behavior change pre- and post-September 11, 2001. The banner ad () linking to the survey was posted in the online chatrooms of a general interest, gay-oriented website between June Crystal methamphetamine (crystal) use in MSM communities 3 and July 24, 2002. Overall, 2284 individuals clicked on the has been problematic in the Western US since the early 1990s banner but exited the survey without answering any questions; [], and has more recently spread to the Midwest [ 3697 clicked on the banner and answered the survey. A total of 2949 questionnaires were complete enough for statistical “marathon sex” (prolonged sexual activity), receptive and analysis (79% completion rate): 2934 were completed by men insertive anal sex without a condom, the ability to have sustained (18 of whom were exclusively heterosexual), 10 by women, arousal for multiple partners, and unsafe sex with and 5 by transgendered individuals. Analysis was limited to the HIV-serodiscordant partners or partners of unknown HIV 2643 men who reported sex with other men or who ,Thus, substance use and its relationship self-identified as gay or bisexual, excluding those who never to high-risk sexual behavior among MSM is of particular had sex, those who were missing information on STDs, and concern, as drugs may help men to avoid feelings of anxiety those who had chronic viral STDs. To assess incident STDs, associated with same-sex behavior and their own awareness of the questionnaire asked if the respondent had been diagnosed with any STDs during two consecutive 3-month periods between This research was undertaken to learn more about risk behaviors June and December 2001 and provided a checklist of the most associated with an incident STD among MSM. We compared common infections. For men reporting viral STDs, we included sexual and drug use behaviors between men with a self-reported only viral STDs that were reported in the second 3-month period.
incident bacterial or viral STD and men without an STD.
Overall, 102 (4%) men reported being diagnosed with anincident bacterial or viral STD during the 6-month study period.
We conducted an anonymous, cross-sectional Internet study,inquiring about sexual and drug-using behaviors among MSM Figure 1. Survey banner ad
In order to minimize non-valid data, we incorporated reliability use before or during sex, alcohol use before or during sex, how checks into the survey for age and certain risk behaviors. To sex partners were met, and HIV testing. Links to STD reduce the likelihood of participants' completing multiple prevention/treatment websites and mental health hotlines surveys, the study banner was rotated through the online chat appeared at the end of the survey. Survey questions were adapted rooms approximately every 20 minutes. Also, it was not from questionnaires used by the investigators in previous studies.
technically possible for participants to bookmark the survey, Data analyses were conducted using SPSS 9.0 for Windows and there were no monetary incentives to complete the survey.
[]. Bivariate categorical data were evaluated using chi-square The general interest, gay-oriented website agreed to host the and odds ratios. Statistically significant bivariate analyses were banner in all of its US adult chat rooms. Individuals had to be simultaneously assessed by multiple logistic regression models.
registered with the website in order to enter chat rooms. The To guard against Type I error, we set the P-value to .01 in the chat room banner provided the only link to the survey. No bivariate analyses, given that controls outnumbered cases almost personally identifying information was collected. The survey did not use cookies and neither collected user IP addresses nor Respondents were asked how many sex partners they had during stored them with submitted data. Study participation was limited two distinct 3-month periods. Respondents could only choose to those 18 and older, and all participants clicked on an online one response from a pull-down menu for each time period.
consent form before gaining access to the anonymous survey.
Answer choices were none, 1, 2-5, 6-10, 11-20, 21-50, and 51 The Medical and Health Research Association of New York or higher. This variable was collapsed for the entire 6-month City, Inc. (MHRA) institutional review board approved the period; men who reported no partners or one were grouped into the first category. Men who reported 2-5 partners were grouped The survey included information on demographics (age group, into the second category, and men who reported 6 partners or race/ethnicity, education, income and residence), and assessment more were grouped into the third. For this analysis, “multiple of risk behaviors, such as type of sexual contact (anal, oral, sex partners” refers to 2 or more partners during the study vaginal; with and without condoms) with main and non-main period. Regarding unprotected anal intercourse (UAI), partners, knowledge of partners’ HIV status, type of illicit drug respondents were asked about insertive and receptive sex J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.2 without a condom. The UAI variable represents men who demographic variables on new registrants from the entire site reported any unprotected receptive and/or insertive anal for the month prior to sample recruitment. Most new registrants intercourse. Age was categorized in a pull-down menu: 18-24, were male (87%), and the study sample was identical to the site 25-29, 30-39, 40-49, 50-59, 60 and older. For ease of analysis, population in age, and similar in educational attainment and by age was collapsed into three categories.
US region. Although two of the demographic variables weresignificantly different, the findings may be a reflection of the large samples sizes, which can detect small differences.
Overall, the study sample was representative of the host websiteuser population. The host website was able to provide several Table 1. Characteristics of study sample and host Internet website
Host Website
Education
US Regional Breakdown
IN, IL, MI, OH, WI, IA, KS, MN, MO, NE, ND, SD DE, DC, FL, GA, MD, NC, SC, VA, WV, AL, KY, MS, TN, AR, LA, OK, TX AZ, CO, ID, NM, MT, UT, NV, WY, AK, CA, HI, OR, WA Chi-square goodness-of-fit statistic used.
Age brackets were recategorized in order to compare to the host website.
Participants resided in all 50 states, roughly in proportion to the Respondents reported newly diagnosed bacterial or viral STDs population of each state. Less than 1% resided in Guam, Puerto (n=102), which included syphilis (n=9), genital herpes (n=4), Rico, and a few locations outside the United States.
genital warts/anal warts/HPV (n=16), gonorrhea (n=49), Approximately half (46%) of the study participants were hepatitis B (n=2), chlamydia (n=29), and non-gonococcal younger than 30 and had at least a college degree (46%). Most urethritis (n=24). Sixteen men reported 2 STDs, and 7 reported were white (85%). Overall, 6% reported crystal use, 7% reported 3 or more. Many respondents in the STD group made an effort cocaine, 9% reported ecstasy, 9% reported Viagra, and about to notify partners of potential exposure. Approximately 30% half (48%) reported drinking alcohol before or during sex. The notified all partners, 26% told some partners but not all, 17% aforementioned drugs were commonly used before or during tried to notify their partners but could not locate them, and less sex (over 85% reported these drugs before or during sex). Most than 5% had the health department notify their sex partners; (80%) reported meeting new sexual partners online, and most (80%) engaged in sex with multiple partners. The majority Men with new STDs were more likely to be between 30 and 39 (81%) engaged in sex exclusively with men, and 7% were years of age than the controls (see ). The bivariate and J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.3 multivariate analyses of risk correlates for STDs were structured or during sex, and sexual risk behaviors (ie, UAI and multiple by drug use and behavioral risk categories as there were no demographic differences between cases and controls. Cases butyrate (GHB), poppers (nitrite inhalants), ketamine, and were significantly more likely to report drug use before or during marijuana use were excluded from the analyses, as their use sex (crystal, cocaine, ecstasy, and Viagra), alcohol use before was not statistically different between groups.
Table 2. Comparison of demographic and behavioral characteristics of men with incident STDs and controls
Controls
Demographics
(n=2541)
Race/Ethnicity
Education
Met Partners Online
HIV Status
* Age, race, and education used logistic regression to calculate the P-value. Income, meeting partners online, and HIV status used chi-square to calculatethe P-value.
To test for multicollinearity, we ran a linear regression with model 2, UAI and having 6 or more sex partners during the “any STD” as the dependent variable and the drug and study were predictive of acquiring an incident STD (UAI OR behavioral risk variables from the bivariate analyses as the 3.9, 95% CI 2.2-7.1, P<.001; 6 or more partners, OR 4.3, 95% independent variables. The variance inflation factor (VIF) value CI 1.8-10.1, P=.001). In the final multiv for each variable was below 1.5, indicating that multicollinearity was not present. We separated the drug and behavioral risk use before or during sex, and having 6 or more sex partners variables for the multivariate logistic analyses into 3 logistic were the strongest predictors of acquiring an incident STD.
regression models to assess risk correlates for acquiring an In order to assess the potential for HIV transmission, we incident STD: model 1 comprised crystal, cocaine, ecstasy, compared the HIV status of the participants to that of their Viagra, and alcohol before or during sex; model 2 comprised partners. Among HIV-positive men with multiple sex partners UAI and number of sex partners during the study; and model 3 who reported UAI (n=109), 47% reported UAI with HIV-negative/unknown partners only, 43% reported UAI with positive and negative/unknown partners, and 10% reported UAI In model 1, only crystal and alcohol before/during sex were with positive partners only. Seven of the 8 respondents with an predictive of acquiring an incident STD (crystal, OR 2.7, 95% STD in this subgroup reported sex with serodiscordant partners.
CI 1.2-6.0, P<.05; alcohol, OR 1.6, 95% CI 1.0-2.6, P<.05). In J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.4 Table 3. Bivariate and multivariate analyses: factors associated with incident STDs
Incident STD
Drug Use and Behavioral Risk
Crystal and Behavioral Risk
Drug Use Before/During Sex
Alcohol Use
Behavioral Risk
Sex partners
*Adjusted odds ratio = the odds ratio estimated after adjusting for all other variables included in the parsimonious model.
Note: In model 1 (data not shown), crystal use and alcohol before/during sex were associated with incident STDs. In model 2 (data not shown), UAIand having 6 or more sex partners were associated with incident STDs.
†Sometimes/most of the time‡Receptive and/or insertive UAIDrug use variables are not mutually exclusive ]. Reback’s report [MSM used crystal to cope with negative internal messages about In this case-control study of men recruited through the Internet, gay sexuality, and HIV-positive MSM reported using it to cope strong associations were found between unprotected anal with the fear of transmitting HIV. Reback’s report also found intercourse, crystal use, and multiple sex partners and an incident that most HIV-positive participants reported that they did not STD. In the overall sample, 4% reported a diagnosis of an disclose their HIV status to casual sex partners as it was their incident bacterial or viral STD during the 6-month study period.
partner’s responsibility to use protection or to set behavioral The great majority of HIV-positive men with multiple sex partners reported unprotected sex with HIV-negative or status Certain limitations of this study deserve mention. Our survey unknown partners, which signifies the continued risk of was posted on only one gay-oriented website. We do not know spreading HIV and other STDs to non-infected indi whether survey respondents would differ if the survey had been Other studies of HIV-positive men report a range of posted on multiple sites or on sites that specifically facilitate serodiscordant or potentially discordant sex, from 21% to 49% meeting sex partners. Minority MSM were underrepresented in []. An average of 80% of our sample met sex partners the sample; our data suggest that white, non-Hispanic MSM online, and study findings indicate risk comparable to other were unintentionally oversampled, as those who have computer Web-based studies on recent sexual risk behavior trends among skills and access to participate in online sex surveys tend to be Men who begin having sex with men while on drugs may self-report only and we did not ask for the site of infection.
develop a pattern of using drugs during sexual experiences [ There may have been underreporting in this sample, as certain and certain drugs such as nitrite inhalants (poppers) and crystal STDs, like chlamydia and gonorrhea, are often asymptomatic may be used specifically to enhance sexual experiences [ Impaired judgment due to drug use may lead to unprotected for certain viral STDs such as genital herpes [, it is not possible to determine whether respondents who participated been hypothesized that substance use may help men avoid in this Internet-based survey are representative of MSM who feelings of anxiety associated with same-sex behavior and use the Internet, of MSM in general, or of MSM with HIV, sincethe MSM population has never been enumerated. Despite these J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.5 limitations, Internet research is an efficient and inexpensive beha]. Primary and secondary substance abuse way to reach large samples of high-risk groups.
treatment among MSM has been successful, as treatment canaffect decisions about sexual behavior uninfluenced by drugs Identifying and treating MSM who have STDs, or who are at and alcohol [wever, treatment remains challenging, and increased risk of acquiring STDs, is crucial in preventing the it has been suggested that men need to abstain from drug use further spread of disease. The Internet is a necessary and and learn skills to meet and initiate sex with men while sober appropriate medium to reach sex-seeking populations for [The rise in crystal use among the MSM population may require a special focus on current substance abuse treatment factoring in increasing numbers of people living with HIV who approaches, such as addressing drugs in exchange for sex, and are resuming sexual activity as a result of improved treatment re]. Just as bathhouses and shooting galleries have integrate HIV/STD and substance use prevention and education been used to deliver STD prevention messages, Internet-based into Web-based and community-based venues. Study findings interventions should be considered for those seeking sex online raise questions concerning the spread of disease and the multiple []. Results of preliminary research on Internet HIV prevention high-risk behaviors, specifically, how drug use is situated within for MSM are promising, suggesting that the Internet may be a the trajectory to unprotected sex, multiple partners, and reliable resource for studying and targeting risk behaviors in ultimately HIV/STD transmission. Additional data are needed to better understand specific pathways between sexual and drug Studies conducted over the past 20 years have found associations using practices and HIV/STD transmission among MSM.
between substance abuse treatment and a reduction in HIV risk Acknowledgements
Data analysis and manuscript preparation were funded in part through CDC Contract Number 200-97-0621, Task 33 to RTIInternational, and Subcontract Number 10-46U-6900 from RTI to Medical and Health Research Association of New York City,Inc. The content of this publication does not necessarily reflect the views or policies of the Department of Health and HumanServices, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.
We would like to thank Mike Humberstone of Data Link, MHRA, Inc. for his technical and programming support.
Conflicts of Interest
References
Centers for Disease Control and Prevention. Increases in HIV diagnoses-29 states, 1999-2002. Morb Mortal Wkly Rep2003;52(47):1145-1148. [ Centers for Disease Control and Prevention. Primary and secondary syphilis among men who have sex with men - NewYork City, 2001. Morb Mortal Wkly Rep 2002;51(38):853-856. [ Rietmeijer CA, Patnaik JL, Judson FN, Douglas JM Jr. Increases in gonorrhea and sexual risk behaviors among men whohave sex with men: a 12-year trend analysis at the Denver metro health clinic. Sex Transm Dis 2003;30(7):562-567. [PMID: Centers for Disease Control and Prevention. HIV prevention through early detection and treatment of other sexuallytransmitted diseases - United States recommendations of the Advisory Committee for HIV and STD Prevention. MorbMortal ] Dickerson MC, Johnston J, Delea TE, White A, Andrews E. The causal role for genital ulcer disease as a risk factor fortransmission of human immunodeficiency virus. An application of the Bradford Hill criteria. Sex Transm Dis1996;23(5):429-440. [PMID: ] Mertz KJ, Weiss JB, Webb RM, Levine WC, Lewis JS, Orle KA, et al. An investigation of genital ulcers in Jackson,Mississippi, with use of a multiplex polymerase chain reaction assay: high prevalence of chanchroid and humanimmunodeficiency virus infection. J Infect Dis 1998;178:1060-1066. [PMID: ] Mbopi-Keou FX, Gresenguet G, Mayaud P, Weiss HA, Gopal R, Matta M, et al. Interactions between herpes simplex virustype 2 and human immunodeficiency virus type 1 infection in African women: opportunities for intervention. J Infect Dis] Erbelding EJ, Chung SE, Kamb ML, Irwin KL, Rompalo AM. New sexually transmitted diseases in HIV-infected patients:markers for ongoing HIV transmission behavior. J Acquir Immune Defic Syndr 2003;33(2):247-252. [PMID: Scheer S, Chu PL, Klausner JD, Katz MH, Schwarcz SK. Effect of highly active antiretroviral therapy on diagnoses ofsexually transmitted diseases in people with ] Colfax GN, Mansergh G, Guzman R, Vittinghoff E, Marks G, Rader M, et al. Drug use and sexual risk behavior amonggay and bisexual men who attend circuit parties: a venue-based comparison. J Acquir Immune Defic Syndr2001;28(4):373-379. [PMID: J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.6 Mansergh G, Colfax GN, Marks G, Rader M, Guzman R, Buchbinder S. The circuit party Men’s Health Survey: findingsand implications for gay and bisexual men. Am J Public Health 2001;91(6):953-958. [PMID: Stall R, Purcell D. Intertwining epidemics: a review of research on substance use among men who have sex with men andits connection to the AIDS epidemic. AIDS and Behavior 2000;4:181-192.
Halkitis P, Parsons J, Stirratt M. A double epidemic: crystal methamphetamine drug use in relation to HIV transmissionamong gay men. J Homose] Semple S, Patterson T, Grant I. Binge use of methamphetamine among HIV-positive men who have sex with men: pilotdata and HIV prevention implications. AIDS Educ Pre] Semple S, Patterson T, Grant I. Motivations associated with methamphetamine use among HIV+ men who have sex withmen. J Subst Abuse T] Halkitis P. The crystal meth-HIV connection. In: HIV Forum NYC New Y [accessed 2004 Nov 20] Reback C, Ditman D. The social construction of a gay drug: methamphetamine use among gay and bisexual males in LosAngeles. Executive Summary. Los Angeles: City of Los Angeles, AIDS Coordinator; 1997.
Whittington W, Collis T, Dithmer-Schreck D, Handsfield HH, Shalit P, Wood RW, et al. Sexually transmitted diseases andhuman immunodeficiency virus-discordant partnerships among men how have sex with men. Clin Infect Dis2002;35(8):1010-1017. [PMID: Mckirnan DJ, Ostrow DG, Hope B. Sex, drugs and escape: a psychological model of HIV-risk sexual behaviours. AIDS McKirnan DJ, Vanable PA, Ostrow DG, Hope B. Expectancies of sexual “escape” and sexual risk among drug andalcohol-involved gay and bisexual men. J Subst Abuse 2001;13(1-2):137-154. [PMID: SPSS Inc. SPSS 1999: Chicago, IL. URL: v 20] Chen SY, Gibson S, Weide D, McFarland W. Unprotected anal intercourse between potentially HIV-serodiscordant menwho have sex with men, San Francisco. J Acquir Immune Def] Halkitis PN, Parsons JT. Intentional unsafe sex (barebacking) among HIV-positive gay men who seek sexual partners onthe Internet. AIDS Care 2003;15(3):367-378. [PMID: Kim AA, Kent C, McFarland W, Klausner JD. Cruising on the Internet highway. J Acquir Immune Defic Syndr2001;28(1):89-93. [PMID: McFarlane M, Bull SS, Rietmeijer CA, The Internet as a newly emerging risk environment for sexually transmitted diseases.
.J] Bull SS, McFarlane M, Rietmeijer C. HIV and sexually transmitted infection risk behaviors among men seeking sex withmen on-line. Am J Public Health 2001;91(6):988-989. [PMID: Lewis L, Ross M. A select body: The gay dance party subculture and the HIV/AIDS pandemic. New York: Cassell; 1995.
Binik Y, Mah K, Kiesler S. Ethical issues in conducting sex research on the Internet. J Sex Res 1999;36(1):82-90.
Lenhart A, Horrigan J, Rainie L, Allen K, Boyce A, Madden M, et al. The ever-shifting Internet population: a new look atInternet access and the digital divide. Washington, DC: Pew Internet & American Life Project. ] Toomey KE, Rothenberg RB. Sex and cyberspace-virtual networks leading to high-risk sex. JAMA 2000;284(4):485-487.
] Turner CF, Rogers SM, Miller HG, Miller WC, Gribble JN, Chromy JR, et al. Untreated gonoccocal and chlamydialinfection in a probability sample of adults. J] Langenberg AG, Corey L, Ashley RL, Leong WP, Straus SE. A prospective study of new infections with herpes simplexvirus type 1 and type 2. Chiron HSV Vaccine Study Group. N Engl J Med 1999;341(19):1432-1438. [PMID: Ross MW, Tikkanen R, Mansson SA. Differences between internet samples and conventional samples of men who havesex with men: implications for research and HIV interv] Rietmeijer CA, Bull SS, McFarlane M. Sex and the Internet. ] Bull SS, McFarlane M, King D. Barriers to STD/HIV prevention on the Internet. Health Educ Res 2001;16(6):661-670.
] Rhodes S, Glorioso J, Hergenrather K. Use of Internet chat rooms for HIV prevention among men who have sex with menin the United States: an evaluation of a community-based initiative. Proceedings of the XIV International Conference onAIDS 2002; Jul 7-12; Barcelona, Spain.
Metzger DS, Navaline H. Human immunodeficiency virus prevention and the potential of drug abuse treatment. Clin Infect] Shoptaw S, Frosch D. Substance abuse treatment as HIV prevention for men who have sex with men. AIDS Behav2000;4(2):193-203.
Abbreviations
HIV: Human immunodeficiency virus
MSM: Men who have sex with men
J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.7 STD: Sexually transmitted disease
UAI: Unprotected anal intercourse
Submitted 17.03.04; peer-reviewed by M McFarlane, G Rebchook; comments to author 25.05.04; revised version received 09.07.04;accepted 19.08.04; published 29.11.04 Please cite as:Hirshfield S, Remien RH, Walavalkar I, Chiasson MACrystal Methamphetamine Use Predicts Incident STD Infection Among Men Who Have Sex With Men Recruited Online: A NestedCase-Control StudyJournal of Medical Internet Research 2004;6(4):e41<URL: J Med Internet Res 2004 | vol. 6 | iss. 4 | e41 | p.8

Source: http://www.betterhealth.is/documents/research/JMIR2004_v6_i4_e41.pdf