Hepatitis C virus infection among short-term intravenous drug users
in southern Taiwan
Chai-Jan Chang
1, Chun-Hui Lin
2, Chien-Te Lee
3, Shun-Jen Chang
4, Ying-Chin Ko
4&
Hong-Wen Liu
11Department of Family Medicine, Kaohsiung Medical College;2Department of Psychiatry, Kai-Suan Psychiatric Hospital; 3Department of Psychiatry, Tsry-Huey Mental Hospital;4School of Public Health, Kaohsiung Medical College
Accepted in revised form 7 June 1999
Abstract. The purpose of this study was to determine the correlation between the prevalence of hepatitis C virus (HCV) infection with duration of drug use and other risk factors among drug users. This sur-vey covered 899 male drug users from Kaohsiung Narcotic Abstention Institute and Kaohsiung prison. The prevalence of positive anti-HCV was 67.2% among intravenous drug users (IVDU) and 14.7% among non-intravenous drug users (non-IVDU). Among intravenous (IV) drug users, age and duration of drug use were independently related to HCV seropositivity. Seroprevalence rate for HCV in
the IVDU group increased with increasing duration of injection use within the ®rst seven years of drug use. However, the steepest trajectory in seropreva-lence of HCV infection occurred within the ®rst four months. Due to the high rate of HCV infection among drug users, investigation of risk behaviors should be routine in such a group. A high frequency of HCV infection was also found among short-term injectors, which indicated that early risk reduction intervention was an important measure in moderat-ing HCV infection.
Key words: Hepatitis C virus infection, Intravenous drug users, Risk factors, Sexual behavior Introduction
In developed countries, intravenous drug users (IVDU) are the only individuals exposed to blood that has not been screened for preventing trans-mission of viral infections. They constitute, therefore, the largest group at risk of viral hepatitis [1±3]. Transmission is primarily parenteral through sharing contaminated hypodermic syringes [4±6].
An important variable to consider when dieren-tiating the risk of HCV infection among drug users is the duration of hypodermic syringe use. A high rate of HCV infection is likely among short-term injectors [7±9]. To be eective, it is important to target pre-ventive intervention at beginners, who are exposed to increased risks. Further studies are required to un-derstand why injectors in certain stages of their drug use are at greater risk of being infected with HCV than in others.
Although very high values of HCV prevalence were reported among Taiwanese drug users [10, 11], so far, few investigations have focused on the association between duration of drug use and HCV infection. The purpose of the present study was to investigate seroprevalence rate in relation to duration of drug use in order to (a) identify the time periods when infection rates were increased and (b) evaluate the
dierences in risk factors among IVDU and non-intravenous (non-IVDU) drug users in Taiwan. Materials and methods
During the period from October 1994 to February 1996, a total of 899 male drug users, aged 14±59 years (mean age: 31.5 years), were examined at the Kaohsiung Narcotic Abstention Institute and Kaohsiung prison. Of these, 247 (27%) had a history of injecting pentazocine and heroin. In this study, an intravenous drug user is de®ned as a person who had ever been an intravenous user of illicit drugs, re-gardless of whether or not he is an illicit intravenous drug user currently.
Anti-HCV testing was done by enzyme immuno-assay (Abbott HCV EIA 2nd generation, Abbott Laboratories, Abbott Park, IL), using a recombinant antigen of viral genome including putative structural and non-structural proteins. All samples had been stored frozen at )70 °C; they were thawed at room temperature and tested according to the manufac-turer's instructions. Specimens with absorbance val-ues greater than the cuto value were considered reactive. The reactive samples were retested and considered positive if both tests were reactive. All
serum samples were tested for HBsAg, anti-HBs and anti-HBc by radioimmunoassay (RIA) using com-mercially available reagents (Abbott Laboratories, Abbott Park, IL). HBV seropositivity was de®ned as a positive ®nding from any of the HBsAg, anti-HBs, or anti-HBc tests.
On the day of blood sampling, copies of a struc-tured questionnaire were distributed among the re-spondents, and all 899 volunteers answered them. The questionnaires included information about de-mographic data, education, occupation, history of hepatitis and blood transfusion, tattoos, age of starting drug use, duration of drug use, average number of injections per day, and occurrence of syringe sharing. Duration of drug use among IVDU subjects is de®ned as the period of time from a participant's reported ®rst injection to his last in-jection of illicit drugs. Although some IVDU sub-jects used illicit drugs by intravenous method intermittently, the duration of drug use was calcu-lated by summing up the time periods of injections. In order to reduce the underreported risk factors, the participants underwent con®dential, face-to-face interviews with a trained family physician in a quiet, private room.
Data were analyzed by SAS 6.04 software package. A v2-test was used for categorical variables to ex-amine the dierence between groups. Odds ratios (OR) with 95% con®dence intervals (CI) were cal-culated to estimate the risks of acquiring HCV in-fection in relation to the risk factors identi®ed by our questionnaire. Those showing signi®cant OR were subsequently examined in a multiple logistic regres-sion model to obtain adjusted OR.
Results
Of the 899 respondents tested, 262 (29.1%) were found positive for anti-HCV (Table 1), and increased with increasing age. The prevalence rate was present in 67.2% of the IVDU group and 14.7% of the non-IVDU group. The results of univariate analyses for postulated risk factors are shown in Table 2. Age was found to be a signi®cant risk factor of HCV infection in the IVDU group. In all, 76% of >30 year old IVDU individuals were found HCV seropositive compared with 47% of O30 year old IV drug users. Histories of blood transfusion, tattooing, and marital status were not risk factors of HCV infection in the IVDU group.
The number of lifetime IVDU sex partners and frequency of condom use were found to be signi®-cantly associated with HCV infection among IVDU, with OR ranging between 3.56 and 4.35 (Table 2). Intravenous drug users with P5 lifetime sex partners were exposed to the same risk as those with <5 lifetime sex partners. A history of gonorrhea or HBV
infection was not signi®cantly related to HCV infec-tion in the IVDU group.
HCV infection was found to be signi®cantly asso-ciated with duration of drug use (Table 2), but not with frequency of drug use or history of needle sharing among IV drug users. The prevalence of HCV infection in IVDU men with >1 month of drug use was signi®cantly greater than in those with O1 month. In the non-IVDU group, these variables (i.e. age, marital status, history of blood transfusion, tattooing, sexual behavior, sexually transmitted dis-ease, duration and frequency of drug use) were not signi®cantly associated with HCV infection (data not shown).
The prevalence of HCV infection in serial cohorts with increasing duration of drug use is depicted in Figure 1, demonstrating a dramatic rise in the prev-alence of HCV infection within the ®rst four months of intravenous drug use, followed by a relatively modest increase in the next seven years. The same prevalence in the non-IVDU group during this period remained essentially unchanged.
We performed logistic regression analysis to de-termine which factors were independently associated with HCV infection in the IVDU group. The sig-ni®cant risk factors of anti-HCV (i.e. age, lifetime IVDU sex partners, frequency of condom use, and duration of drug use) revealed in univariate analysis were included into the logistic regression analysis (Table 2). Age and duration of IV drug use were statistically signi®cantly and independently associ-ated with HCV infection. HCV seropositivity was found to be signi®cantly associated with IVDU men older than 30 years old, who injected drugs for longer than one month (odds ratio, 2.96 and 5.14). The number of IVDU sex partners and frequency of condom use were not found associated with HCV infection according to logistic regression modeling.
Table 1. Prevalence of anti-hepatitis C virus by type of drug use and age among drug users
Characteristics No. tested No. (%) positive p Overall 899 262 (29) Type of drug use
Intravenous 247 166 (67) <0.001 Other 652 96 (15) Age, years <20 29 4 (14) <0.001 20±24 121 29 (24) 25±29 248 44 (18) 30±34 228 63 (28) 35±39 142 65 (46) 40±44 92 40 (43) P45 39 17 (44)
Discussion
The data presented here showed that HCV infection was common among drug users. The prevalence of positive anti-HCV was 67.2% in the IVDU group and 14.7% in the non-IVDU group. The seropreva-lence in the IVDU men in southern Taiwan (67%) was somewhat lower than in similar groups of people in Spain (73%), Amsterdam (74%), Baltimore (85%), and comparable to those in Rome (67%) and New Zealand (67%) [4, 5, 7, 12, 13].
A major ®nding of this study was the dramatic rise of HCV infection observed within the ®rst 7 years of intravenous drug use. Seroprevalence of HCV
(86.7%) in the cohort embracing men with a history of 49 to 72 months of injection was slightly higher than that observed in the ALIVE study (85.0%) [9]. We also found that the seroprevalence rate increased in these seven years with increasing duration of in-jection use. The steep trajectory in seroprevalence within the ®rst four months of injection use suggested that most new HCV infections occurred soon after beginning intravenous drug use. Prevalence of HCV infection in the 2 to 4 month cohort was signi®cantly higher than that in the O1 month group. There are several obvious explanations for this result. In the ®rst month of IV drug use, injections are relatively infrequent and this reduces the risk of infection.
Table 2. Analysis of risk factors associated with HCV among IVDU subjects
Variable N N (%) Positive OR (95% CI) aOR (a95% CI) Age, years >30 169 129 (76.3) 3.57 (2.04±6.20) 2.96 (1.64±5.36) O30 78 37 (47.4) 1.00 1.00 Marital status Single 169 116 (68.6) 1.23 (0.69±2.19) ± Married 78 50 (64.1) 1.00 Blood transfusion Yes 37 29 (78.4) 1.93 (0.85±4.39) ± No 210 137 (65.2) 1.00 Tattooing Yes 161 113 (70.2) 1.47 (0.84±2.56) ± No 86 53 (61.6) 1.00
Lifetime sex partners
P5 52 35 (67.3) 1.00 ±
<5 195 131 (67.2) 1.00
Lifetime IVDU sex partners
P2 23 20 (87.0) 3.56 (1.09±11.60) 2.62 (0.73±9.35) <2 224 146 (65.2) 1.00 1.00 Condom use Never/seldom 238 163 (68.5) 4.35 (1.18±16.10) 2.97 (0.66±13.27) Always 9 3 (33.3) 1.00 1.00 Gonorrhea Yes 62 36 (58.1) 0.59 (0.32±1.06) ± No 185 130 (70.3) 1.00 HBV Positive 239 163 (68.2) 3.57 (0.90±14.11) ± Negative 8 3 (37.5) 1.00
Drug use frequency
P4 times daily 106 76 (71.7) 1.44 (0.83±2.48) ± <4 times daily 141 90 (63.8) 1.00
Duration of drug use, months
>1 231 162 (70.1) 7.04 (2.51±19.77) 5.14 (1.53±17.27)
O1 16 4 (25.0) 1.00 1.00
Needle sharing
Yes 73 52 (71.2) 1.30 (0.72±2.35) ±
No 174 114 (65.5) 1.00
Anti-HCV indicates antibody to hepatitis C virus; IVDU, intravenous drug user; OR, odds ratio; CI, con®dence interval; aOR, adjusted odds ratio; a95% CI, adjusted 95% con®dence interval.
Furthermore, a number of factors typical of drug addiction may contribute to HCV infection, such as polydrug abuse, tattooing, and iatrogenic injections.
Univariate analyses indicated that age, duration of drug use, number of lifetime IVDU sex partners, and frequency of condom use were all signi®cantly asso-ciated with HCV infection among the IV drug users. However, age and duration of drug use were inde-pendently related to the rate of seropositive HCV infection, suggesting that HCV was readily trans-mitted parenterally. This ®nding is comparable to other studies [4±8]. Giradi et al. [4] found that anti-HCV presence was associated with duration of IV drug use. Donahue et al. [5] documented a signi®cant association between duration of IV drug use and hepatitis C seropositivity. History of needle sharing was not signi®cantly associated with HCV infection among the IVDU group. The reason for this may be explained by the fact that most of our subjects might have underreported their risk behaviors in the prison environment.
Although the prevalence of HCV infection in the non-IVDU group was much lower than that in IV drug users, it was still 10 times higher than that (1.5%) observed among 20,768 healthy blood donors in Taiwan [14]. This dierence could not be explained by age distribution, marital status, tattooing, history of blood transfusion, sexual behavior, or sexually transmitted diseases. However, a high frequency of non-needle sharing subjects with HCV infection in the IVDU group was found in our study (66%), and age was an independent risk factor for HCV infec-tion. It indicated other routes of transmission which we had not investigated, such as that reported by Chang et al. [15]. In their study, they found that iatrogenic injections, frequently given at illegal medical service stations, played an important role in HCV infection in a Taiwanese aboriginal community.
In contrast with our present ®ndings, several studies have found that HCV may be sexually transmitted [6, 16±18]. As age is a cumulative factor related signi®cantly to sexual behavior, it may de-crease any association between sexual behavior and HCV infection. According to univariate analysis in our study, however, age was an independent risk factor of HCV infection, which was signi®cantly as-sociated with sexual behavior among the IV drug users. This would suggest that sexual transmission may still occur, albeit infrequently.
Respondents may have inaccurately answered or underreported certain risk behaviors, such as intra-venous drug use, sexual habits, syringe sharing, etc., and this may have limited the value of the present ®ndings. In addition, we may have underestimated the seroprevalence of HCV in our population sample, due to seroconversion and the fact that the second-generation assay used here may have failed to identify patients either with low titers of anti-HCV or with antibodies to other antigens of HCV [19]. The small number of each age group (Figure 1) may also have limited the power of our study in identifying the changeable prevalence of HCV infection.
These ®ndings suggest that intravenous drug use is the predominant risk factor of HCV infection. These data also strongly suggest that the interval soon after beginning injections is a particularly dangerous pe-riod for infection with the hepatitis C virus. Due to the high levels of infection within the ®rst four months of drug use, our data provide an additional rationale to reemphasize the need for primary sub-stance abuse prevention. The prevalence of HCV in-fection among drug users is very much higher than among blood donors in Taiwan. However, it would be worthwhile to carry out further studies into this serious public health problem.
Acknowledgements
The study was supported by grant DOH 85-TD-128 from the Department of Health, Executive Yuan, Republic of China.
References
1. Patti AM, Santi AL, Pompa MG, et al. Viral hepatitis and drugs: A continuing problem. Int J Epidemiol 1993; 22: 135±139.
2. Norkrans G, FroÈsner G, Hermodsson S, Iwarson S. Multiple hepatitis attacks in drug addicts. JAMA 1980; 243: 1056±1058.
3. Wright R. Viral hepatitis comparative epidemiology. Brit Med Bull 1990; 46: 548±558.
4. Girardi E, Zaccarelli M, Tossini G, et al. Hepatitis C virus infection in intravenous drug users: Prevalence and risk factors. Scand J Infect Dis 1990; 22: 751±752. Figure 1. The prevalence of hepatitis C virus infection in
serial cohorts with increasing duration of drug use among IVDU and non-IVDU, not including the cases with more than 7 years of drug use.
5. Donahue JG, Nelson KE, Munoz A, et al. Antibody to hepatitis C virus among cardiac surgery patients, ho-mosexual men, and intravenous drug users in Baltimore, Maryland. Am J Epidemiol 1991; 134: 1206±1211. 6. Osmond DH, Charlebois E, Sheppard HW, et al.
Comparison of risk factors for hepatitis C and hepatitis B virus infection in homosexual men. J Infect Dis 1993; 167: 66±71.
7. van den Hock JAR, van Haastrecht HJA, Goudsmit J, de Wolf F, Coutinho RA. Prevalence, incidence and risk factors of hepatitis C virus infection among drug users in Amsterdam. J Infect Dis 1990; 162: 823±826. 8. Bell J, Batey RG, Farrel GC, et al. Hepatitis C virus in
intravenous drug users. Med J Aust 1990; 153: 274±276. 9. Garfein RS, Vlahov D, Galai N, Doherty MC, Nelson KE. Viral hepatitis in short-term injection drug users: The prevalence of the hepatitis C, hepatitis B, human immunode®ciency, and human T-lymphotropic viruses. Am J Public Health 1996; 86: 655±661.
10. Lee SD, Chan CY, Wang YJ, et al. Seroepidemiology of hepatitis C virus infection in Taiwan. Hepatology 1990; 13: 830±833.
11. Chen DS, Wang JT, Chen PJ, Wang TH, Sung JL. Hepatitis C virus infection in Taiwan. Gastroenterol Jpn 1991; 26 (suppl 3): 164±166.
12. Esteban JI, Viladomiu L, Gonzalez A, et al. Hepatitis C virus antibodies among risk groups in Spain. Lancet 1989; 8: 294±296.
13. Wood®eld DG. Hepatitis C virus (HCV) infections in New Zealand. Gastroenterol Jpn 1991; 26 (suppl 3): 189±191.
14. Wang JT, Wang TH, Sheu JC, et al. Hepatitis C virus infection in volunteer blood donors in Taiwan:
Evalu-ation by hepatitis C antibody assays and the polymer-ase chain reaction. Arch Pathol Lab Med 1993; 117: 152±156.
15. Chang SJ, Chen HC, Ying J, Lu CF, Ko YC. Risk factors of hepatitis C virus infection in a Taiwanese aboriginal community. Kaohsiung J Med Sci 1996; 12: 241±247.
16. Alter MJ, Coleman PJ, Alexander WJ, et al. Impor-tance of heterosexual activity in the transmission of hepatitis B and non-A, non-B hepatitis. JAMA 1989; 262: 1201±1205.
17. Thomas DL, Cannon RO, Shapiro CN, et al. Hepatitis C, hepatitis B, and human immunode®ciency virus in-fections among non-intravenous drug-using patients attending clinics for sexually transmitted diseases. J Infect Dis 1994; 169: 990±995.
18. Nakashima K, Kashiwagi S, Hayashi J, et al. Sexual transmission of hepatitis C virus among female pros-titutes and patients with sexually transmitted diseases in Fukuoka, Kyushu, Japan. Am J Epidemiol 1992; 136: 1132±1137.
19. Alter HJ, Purcell RH, Shih JW, et al. Detection of antibody to hepatitis C virus in prospectively fol-lowed transfusion recipients with acute and chronic non-A, non-B hepatitis. N Engl J Med 1989; 321: 1494±1500.
Address for correspondence: Chai-Jan Chang, Department of Family Medicine, No 100 Shih-Chuan 1st Road, Kaohsiung City 807, Taiwan, Republic of China
