Abstract
HIV risk perception remains an effective determinant of HIV transmission. Although higher educational attainment has been associated with increased HIV risk perception, this predictor remains to be assessed among Nigerian military personnel (NMP). In a prospective cohort of 2,213 NMP, the effect of education and other factors on HIV risk perception were assessed at baseline using chi square statistic and unconditional logistic regression. There was an inverse correlation between higher educational attainment and HIV risk perception in the univarible model, prevalence odds ratio (POR), 0.64, 95% confidence interval (CI) = 0.52–0.79. This association persisted after adjustment for relevant covariates in the multivariable model (POR, 0.70, 95% CI=0.56–0.88). Likewise, there was a direct correlation between use of alcohol and marijuana and HIV risk perception (p <0.05). In contrast, casual sex and gender were not statistically significantly associated with HIV risk perception, P >0.05. This study is indicative of an inverse correlation between educational attainment and HIV risk perception, as well as a direct correlation between alcohol and marijuana and HIV risk perception among NMP. Therefore HIV prevention interventions targeted at NMP need to include multiple factors that may impact on risk perception regardless of educational status of the participants.
Keywords: HIV risk perception, educational status, military personnel, cross-sectional design, prevalence odds ratio
INTRODUCTION
Nigeria is the second most affected country in sub-Saharan Africa (SSA) with HIV, representing 14% of HIV/AIDS cases in this region.1 While the first HIV case was reported in Nigeria in 1986, the HIV infection prevalence proportion increased from 1.8% in 1991 to 4.5% in 1996 to 5.8% in 2001, with a slight decrease to 5.0% in 2003.2 This prevalence varied by occupation and across state boundaries with previous studies of 2300 subjects from five states in Nigeria demonstrating a prevalence of over 60% among commercial sex workers (CSW), 8% among male clients of CSW, 9% among truck drivers, and 21% among patients with sexually transmitted infections.3 Among Nigerian military personnel (NMP), HIV prevalence has been reported to be higher than that of the general population,4 due to the mobile lifestyle and distance from their spouses while on United Nations peace-keeping mission.5,6,7–9
There are variables associated with HIV risk perception that have been studied in the general population, but have yet to be fully investigated among Nigerian military personnel (NMP). Examples include the association between HIV risk perception and educational attainment as well as the relationship between HIV risk perception and knowledge of HIV risk factors.10,11 An epidemiologic study has shown that conventional HIV risk such as multiple sex partners, drug use and inconsistent condom use, compared with HIV risk perception is a lesser determinant of HIV infection.12 The association between higher educational attainment and increased knowledge of HIV risk factors as well as education and HIV risk perception has been well documented across studies in most developed countries, but not among military personnel in sub-Saharan Africa.13,14,15
Behavioral and epidemiologic factors leading to HIV infection and AIDS morbidity remain to be fully explored, particularly among mobile populations who are at heightened risk of HIV infection. Increasing HIV risk perception may enhance HIV-related protective behaviors, thus decreasing the propensity for HIV infection.16–18. In this sample of NMP, HIV risk perception may be influenced by multiple factors including educational attainment, subjective norms, actual risk, income, and marital status. Whereas the nexus between HIV risk perception and educational attainment has been studied in United States populations and other civilian-based populations, to our knowledge there is no study that has examined the influence of educational status on individual’s risk perception of HIV infection among NMP. We hypothesized that among NMP, higher education may be associated with HIV risk perception and that higher educational attainment directly correlates with HIV risk perception. This study therefore represents the first published cross-sectional investigation on educational attainment as a predictor of HIV risk perception among Nigerian military personnel.
MATERIALS AND METHODS
STUDY PARTICIPANTS
Study participants were Nigerian military personnel, namely army, navy and air force. The study sample consisted of a cohort of 2,213 men and women, aged 18 to 55, recruited in 2003 for an HIV education intervention. Of the 2,213, 13.3% were women and 86.7% were men, 43.9% had less than high school education and 56.1% had high school and some college education. The details of the materials and methods are available elsewhere.5, 6, 19
STUDY DESIGN
A cross-sectional design was utilized to assess the association between HIV risk perception as the outcome variable and educational status as well as other potential predictors. Two cantonments stationed in the Lagos area, the largest city in Africa, were selected for intervention. Data were collected between June and December 2003.
A modified version of the Center for AIDS Intervention Research at the University of Wisconsin, Medical College (Department of Psychiatry and Behavioral Medicine) Pridefest Survey was used. This instrument obtains information on socio-demographics (sex, age, ethnic background, educational attainment, HIV status, gender of sexual partners, having main sexual partner), sexual activities (sex with main or casual partners, frequency of vaginal, oral or anal sex with or without condom), drug use (alcohol, marijuana, amphetamines, ecstasy, cocaine, heroin), and beliefs about condoms.
VARIABLES MEASURE
Outcome/Response variable
The outcome variable in this study was HIV risk perception. This variable was measured using a seven-item scale. Examples of the questions in this construct included: “When I am high or drunk, I am more likely to get into a situation that may lead to sex”; “When I am high or drunk, I am more likely to use a condom”; “I am worried about getting AIDS”; and “I will use condoms in the heat of the moment”. This construct (HIV risk perception) had an average inter-item covariance of 0.05 and a very strong reliability score of 0.89 Cronbach α. The items in the construct were measured on a scale of 1 to 10 with 1 indicating the greatest level of agreement implying lack (negative) of HIV risk perception and 10 indicating the highest level of disagreement, implying the presence (positive) of HIV risk perception. Where the wording of the question was reversed, the scoring of the items was equally reversed to correlate with negative or positive HIV risk perception.
Predictor/Independent variables
Education attainment
The primary predictor or independent variable for this study was educational attainment. This variable was measured using a categorical scale subdivided into four classes: (1) elementary, (2) junior secondary, (3) senior secondary and (4) some college. To present the findings in a way that could be understood by researchers across geographic boundaries, the United States category was used for the equivalence of educational attainment and generated two comparable categories: (1)less than high school and (2) equal to and greater than high school. Less than high school is comparable to elementary and junior secondary category while equal to and greater than high school is comparable to senior secondary and some college education.
Other potential predictors: age, gender, casual sex practice, alcohol and marijuana use, and barrier to condom acquisition
The other independent variables as potential predictors of HIV risk perception were age, gender, sexual orientation, casual sex practice, alcohol and marijuana use, and barriers to condom use. A categorical scale was used to measure age, alcohol and marijuana use while a binary scale was used to measure gender and casual sex. Alcohol and marijuana use were measured as the frequency of use of these substances during the past six weeks. The information was collected in three categories: none (zero), 1 to 4 times, and (3) greater than 4 times. Barrier to condom acquisition was a construct based on a two-item scale. The items in the scale were: (1) “It will be embarrassing to buy condoms in a store” and(2) “I will feel uncomfortable carrying condoms with me.” The average inter-item covariance for this construct was 0.05, while the Cronbach α for the scale reliability was 0.70.
Human Subjects and Ethical Consideration
The research protocol was approved by relevant institutional review boards (IRB) in the United States and Nigeria. All participants provided informed consent.
STATISTICAL ANALYSIS
Chi square statistic with Fisher’s exact (correcting for small cells count) was used to assess differences in baseline characteristics (age, gender, ethnicity, casual sex practice, alcohol and marijuana use) of study participants by educational status. Using unconditional univariable logistic regression model, we examined separately the relationship between HIV risk perception and the following variables: participants’ age, gender, educational attainment, alcohol and marijuana use, ethnicity as membership in tribal groups, casual sex practice, and barriers to condom use. Next, we performed multivariable analysis using unconditional logistic regression to simultaneously control for the possible confounding effects of these variables on the influence of educational status on HIV risk perception. Further, we examined the statistical interaction of age and ethnicity and ethnicity and education at the significance level, p < 0.10.20 We determined a priori that the independent variables in the final multivariable logistic model would be those significantly associated with HIV risk perception in the univariable analyses, using a statistically significant level, p < 0.25.20 We entered into the final model variables that were statistically significant at p < 0.25 and biologically relevant and performed the Goodness of Fit test using Hosmer-Lemeshow criteria,20 and regression diagnostic to see if the model fit. All statistical analyses were two-tailed at 0.05 significance level and were performed using STATA statistical package, version 9.2.
RESULTS
Table I presents the comparison of study characteristics by educational status. The sample comprised a total of 2,213 subjects, with the majority being men, 1,918 (86.7%). Although not shown on the table, a small proportion of study participants perceived that they were at risk for HIV infection (21.4%), 43.9% had less than high school education, and 56.1% had high school or some college education. Women were more likely to perceive themselves at risk for HIV infection compared to men, (27.3% versus 20.5%, respectively). There were no statistically significant differences in gender, casual sex, and alcohol use with respect to educational status. However, there were statistically significant differences in marijuana use and barriers to condom acquisition (p < 0.05).
Table I.
The comparison of study characteristics and predictive factors by educational status
| Educational Status
|
|||||||
|---|---|---|---|---|---|---|---|
| < High School
|
≥ High School
|
||||||
| Covariates | Number (n) | Percentage (%) | Number (n) | Percentage (%) | χ2 | df | p-value |
| Gender | 0.1 | 1 | 0.74 | ||||
| Female | 127 | 13.1 | 168 | 13.5 | |||
| Male | 845 | 86.9 | 1072 | 86.5 | |||
| Age (Year) | 68.1. | 2 | < 0.001 | ||||
| 18 – 29 | 581 | 59.8 | 636 | 51.3 | |||
| 30 – 49 | 327 | 33.6 | 584 | 47.1 | |||
| ≥ 50 | 64 | 6.6 | 19 | 1.5 | |||
| Casual sex | 0.0 | 1 | 0.99 | ||||
| Yes | 480 | 49.7 | 611 | 49.7 | |||
| No | 485 | 50.3 | 618 | 50.3 | |||
| Alcohol use | 1.9 | 2 | 0.40 | ||||
| none (0) | 419 | 43.2 | 533 | 43.2 | |||
| 1 – 4 | 436 | 45.0 | 577 | 46.8 | |||
| ≥ 5 | 114 | 11.8 | 124 | 10.1 | |||
| THC use | 34.2 | 2 | < 0.001 | ||||
| none (0) | 693 | 71.5 | 991 | 80.3 | |||
| 1 – 4 | 246 | 25.4 | 191 | 15.5 | |||
| ≥ 5 | 30 | 3.1 | 52 | 4.2 | |||
| Barrier to Condom acquisition | 31. 2 | 1 | < 0.001 | ||||
| Yes | 762 | 78.8 | 1073 | 87.7 | |||
| No | 205 | 21.2 | 151 | 12.3 | |||
Abbreviations: degrees of freedom=df, chi square= χ2, tetrahydrocanabinnol (THC) = marijuana.
Table II presents the univariable unconditional logistic regression analysis of the association between HIV risk perception and educational attainment as well as other socio-demographics and predictor variables. Compared with females, males were 32.0% less likely to perceive themselves at risk for HIV infection (POR = 0.68, 95% CI = 0.51 – 0.91). Relative to the youngest age group (18 – 29 years), the intermediate age group (30 – 49 years) was 46.0% less likely to perceive themselves at risk for HIV infection (POR = 0.54, 95% CI = 0.43 – 0.67). There was a statistically significant difference in HIV risk perception among those with less than high school education compared to participants with high school education and some college. Compared to subjects with less than high school, those with high school and some college were 36.0% less likely to perceive themselves at risk of HIV infection (POR = 0.64, 95% CI = 0.52 – 0.79). Also, compared to participants who reported having casual sex, those who reported no involvement in casual sex were 26.0% less likely to perceive themselves at risk for HIV infection (POR = 0.74, 95% CI = 0.60 – 0.91). Table II also presents the comparison of subjects who reported not using marijuana or alcohol to those who reported marijuana or alcohol use 1 to 4 times during the past 6 weeks. Compared to those not using marijuana, those using marijuana 1 to 4 times during the past six weeks were 81.0% more likely to perceive themselves at risk for HIV infection (POR = 1.81, 95% CI = 1.42 – 2.30). Likewise, alcohol consumption was statistically significantly associated with HIV risk perception. Compared to participants who reported no alcohol use during the past 6 weeks, those who reported use of alcohol 1 to 4 times during the same time frame were 92.0% more likely to perceive themselves at risk for HIV infection (POR = 1.92, 95% CI = 1.53 – 5.41).
Table II.
The Univariable unconditional logistic regression model of the association between HIV risk perception and education and other predictors among Military Personnel
| Covariates | HIV Risk Perception (Number)
|
POR | 95% Confidence Interval | |
|---|---|---|---|---|
| HRP (−ve) | HRP(+ve) | |||
| Gender | ||||
| Female | 202 | 76 | 1.0 | reference |
| Male | 1457 | 376 | 0.68 | 0.51 – 0.91 |
| Age Group | ||||
| 18 – 29 | 866 | 302 | 1.0 | reference |
| 30 – 49 | 729 | 136 | 0.54 | 0.43 – 0.67 |
| >50 | 64 | 15 | 0.67 | 0.38 – 1.20 |
| Educational status | ||||
| < High School | 698 | 239 | 1.0 | reference |
| ≥ High School | 961 | 212 | 0.64 | 0.52 – 0.79 |
| Casual Sex | ||||
| Yes | 802 | 252 | 1.0 | reference |
| No | 846 | 197 | 0.74 | 0.60 – 0.91 |
| Alcohol use | ||||
| None | 761 | 144 | 1.0 | reference |
| 1–4 times | 716 | 260 | 1.92 | 1.53 – 2.41 |
| > 5 times | 176 | 48 | 1.49 | 1.00 – 2.08 |
| Marijuana use | ||||
| None | 1294 | 311 | 1.0 | reference |
| 1–4 times | 292 | 127 | 1.81 | 1.42 – 2.30 |
| > 5 times | 67 | 14 | 0.87 | 0.48 – 1.57 |
| Barrier to condom acquisition | ||||
| Yes | 144 | 1932 | 1.0 | reference |
| No | 307 | 26 | 133.8 | 86.61 – 206.7 |
Abbreviations: HIV risk perception=HRP, Presence of HRP =+ve, Absence of HRP=−ve, unadjusted prevalence odds ratio=POR
Notes: Marijuana and alcohol use refers to the frequency of use during the past six weeks.
Table III presents the multivariable unconditional logistic regression model for the association between educational status and HIV risk perception, adjusting for age, gender, casual sex, alcohol use, and marijuana use. After controlling for these covariates, education was still statistically significantly associated with HIV risk perception. Compared to participants with less than high school education, those with high school and some college education were 30.0% less likely to perceive themselves at risk of HIV infection ( adjusted POR (APOR) = 0.70, 95% CI = 0.58 – 0.88). Relative to the youngest age group (18 – 29 years), those in the intermediate age group (30 – 49 years) were 43.0% less likely to perceive themselves at risk of HIV infection (APOR = 0.57, 95% CI = 0.45 – 0.73). After adjustment for these covariates, there was no statistically significant difference in HIV risk perception by gender and casual sex. Also, compared with participants who reported not using alcohol during the past 6 weeks, those who reported using alcohol 1 to 4 times in the past 6 weeks were almost 3 times more likely to perceive themselves at risk for HIV infection (APOR = 2.85, 95% CI = 2.17 – 2.74). Compared to those who reported not using marijuana, those who reported using marijuana 1 to 4 times during the past 6 weeks were 3 times more likely to perceive themselves at risk of HIV infection (APOR = 3.08, 95% CI = 2.30 – 4.12).
Table III.
Multivariable logistic regression model for the association between educational status and HIV risk perception
| Covariate | (APOR) | 95% CI | p-value |
|---|---|---|---|
| Education | |||
| < High Sch | 1.0 (reference) | reference | reference |
| ≥ High Sch | 0.70 | 0.56 – 0.88 | < 0.001 |
| Age Group | |||
| 18 – 29 | 1.0 (reference) | reference | reference |
| 30 – 49 | 0.57 | 0.45 – 0.73 | < 0.001 |
| ≥ 50 | 0.65 | 0.36 – 1.19 | 0.16 |
| Gender | |||
| Female | 1.0 (reference) | reference | reference |
| Male | 2.21 | 0.71 – 6.85 | 0.17 |
| Casual Sex | |||
| Yes | 1.0 (reference) | reference | reference |
| No | 0.91 | 0.72 – 1.14 | 0.41 |
| Alcohol use | |||
| Zero | 1.0 (reference) | reference | reference |
| 1 – 4 | 2.85 | 2.17 – 3.74 | < 0.001 |
| ≥ 5 | 2.30 | 1.53 – 3.46 | < 0.001 |
| Marijuana use | |||
| Zero | 1.0 (reference) | reference | reference |
| 1 – 4 | 3.08 | 2.30 – 4.12 | < 0.001 |
| ≥ 5 | 1.72 | 0.89 – 3.31 | 0.10 |
Abbreviations: Adjusted prevalence odds ratio (APOR), Sch=School, CI = Confidence Interval.
Notes: Marijuana and alcohol use refers to the frequency of use during the past six weeks.
DISCUSSION
There are some relevant findings from this study. First, among military personnel, educational attainment inversely correlates with HIV risk perception. Secondly, alcohol and marijuana use significantly increase HIV risk perception in this cohort. Thirdly, there was an inverse correlation between age of participants and HIV risk perception. An additional potential relevant finding, though not statistically significant, was a direct correlation between casual sex and HIV risk perception.
In this study, we have shown that among NMP, higher education is associated with decreased HIV risk perception. Our finding in this vein is not supported by most findings regarding the association between higher education and HIV risk perception. In the general or civilian population, where there is a relative increase in exposure to HIV information, studies have found a significant increase in HIV risk perception among participants with higher education.13,15 There is a plausibility in our finding that higher education in this sample inversely correlates with HIV risk perception. Therefore, despite higher educational attainment, this cohort may not be exposed to substantial and reliable information on risk factors for acquiring HIV infection. In this cohort of NMP, educational attainment may not necessarily translate to knowledge of HIV risk factors, which is necessary for increased risk perception, and higher education status may be associated with higher rank among NMP, which may precipitate denial syndrome.
In addition, we have demonstrated that alcohol and marijuana use significantly increase HIV risk perception in this cohort. The finding of direct correlation between alcohol and marijuana use during the past six weeks and HIV risk perception simply suggests a direct correlation between a perceived risk and actual risk factors for HIV infection. Substance use, including alcohol and marijuana, are actual risk factors for HIV infection and the drug addicted population is associated with increased prevalence of HIV infection.21–24 The finding in this study supports a previous study conducted among a civilian population, which observed increased risk for HIV infection among drug users.25 Our finding is plausible in the sense that exposure to drug use may influence the individual’s perception of the risk of HIV infection.
The data also showed a direct association between casual sex and HIV risk perception. However, this finding was not statistically significant. It is probable that decreased HIV risk perception may lead to modification in casual sex engagement. This finding is not surprising given the nature of our data, cross-sectional, that precludes an assessment of temporality as to which behavior preceded the other, HIV risk perception or casual sex engagement.
Furthermore, we have illustrated an inverse correlation between the age of participants and HIV risk perception. This finding, though unanticipated, is conceivable in this cohort. There is a possibility that advancing age is associated with decreased vulnerability to a disease that is associated with sexual intercourse since sexuality is perceived to decrease with aging. Secondly, HIV prevention messages in Nigeria traditionally target younger populations and, hence, older adults in our cohort may not have been exposed to such information which could potentially decrease their perception of HIV risk.
There are some limitations to our findings. First, as a cross-sectional data, it is difficult to establish a temporal sequence between outcome and independent variables. Secondly, like in all epidemiologic or observational studies, our finding may be influenced by residual or unmeasured confounding. However, it is unlikely that our finding of association between educational attainment and HIV risk perception is solely dependent on the confounding effects of age, gender, casual sex practice, alcohol and marijuana use. Thirdly, the point estimate for the effect of educational attainment on HIV risk perception may be influenced by misclassification or selection bias. Misclassification bias is possible in that we re-categorized and transformed variables prior to analysis. These transformations have tendencies toward introducing non-differential misclassification and selection bias. However, misclassification bias is highly unlikely in this sample since we examined categories for the stability of the prevalence odds ratio and collapsed categories when and where necessary.
In summary, we have shown that in this cohort of Nigerian military personnel, higher educational attainment inversely correlates with HIV risk perception. In addition, alcohol and marijuana use are associated with HIV risk perception. While this study suggests increasing HIV risk perception through information on risk and focus on situational variables such as drug and alcohol use, further studies are needed to explore whether the perception of HIV risk by marijuana and alcohol use results in HIV risk reduction.
Acknowledgments
This study was funded by grant number (01-038-258) from the World AIDS Foundation, Paris, France. Preparation of this manuscript was facilitated by National Institute of Mental Health grant number RO1 MH073361-02. The authors thank Jennifer Krueger and Jonathan Brunt for editing the initial draft of this manuscript.
Contributor Information
E. James Essien, 1441 Moursund Street, Houston, TX 77030.
Gbadebo O. Ogungbade, 1441 Moursund Street, Houston, TX 77030.
Doriel Ward, 1515 Holcombe, Houston, TX 77030.
Ernest Ekong, 250A Ikorodu Crescent, Ikoyi, Lagos, Nigeria.
Michael W. Ross, Center for Health Promotion and Prevention Research, P.O.Box 20036, Houston, TX 77225.
Angela Meshack, Center for Health Promotion and Prevention Research, P.O.Box 20036, Houston, TX 77225.
Laurens Holmes, Jr, 1441 Moursund Street, Houston, TX 77030.
References
- 1.UNAIDS, 2004 – 2004 Report on the Global AIDS Epidemic, July, 2004.
- 2.UNAIDS, 2005 – 2005 Report on the Global AIDS Epidemic, July, 2005.
- 3.Esu-Williams E, Mulanga-Kabeya C, Takena H, Zwandor A, Aminu K. Seroprevalence of HIV, HIV-2, and HIV-1 group O in Nigeria: Evidence of an increase of HIV infection. Journal of Acquired Immune Deficiency Syndrome. 1997;16:204–210. doi: 10.1097/00042560-199711010-00010. [DOI] [PubMed] [Google Scholar]
- 4.Nwokoji UA, Ajuwon AJ. Knowledge of AIDS and HIV risk –related sexual behavior among Nigerian Naval Personnel. BMJ Public Health. 2004;4:24–32. doi: 10.1186/1471-2458-4-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Essien EJ, Meshack AF, Ekong E, et al. Effectiveness of a situationally-based HIV risk-reduction intervention for the NigerianUniformed Services on readiness to adopt condom use with casual partners. Counselling, Psychotherapy, and Health. 2005;1:19–30. [Google Scholar]
- 6.Essien EJ, Ogungbade GO, Holmes L, et al. Emerging socio-demographic and lifestyle predictors of intention to use condom in human immunodeficiency virus intervention among Uniformed Services Personnel. Military Medicine. 2006;171:1027–1034. doi: 10.7205/milmed.171.10.1027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Miles SH. HIV in insurgency forces in sub-Saharan Africa – a personal view of policies. International Journal of STD and AIDS. 2003;14:174–178. doi: 10.1258/095646203762869179. [DOI] [PubMed] [Google Scholar]
- 8.Fleshman M. AIDS prevention in the ranks: UM targets peacekeepers, combatants in war against disease. Africa Recovery. 2001;15:1–2. [Google Scholar]
- 9.Newman LM, Miguel F, Jemuse BB. HIV seroprevalence among military blood donors in Manica Province, Mozambique. International Journal of STDs and AIDS. 2001;12:225–228. doi: 10.1258/0956462011922968. [DOI] [PubMed] [Google Scholar]
- 10.Harrison DF, Wambach KG, Byers JB, Imershein AW, Levine P, Maddox K, Quadagno DM, Fordyce ML, Jones MA. AIDS knowledge and risk behaviors among culturally diverse women. AIDS Educ Prev. 1991;3:79–89. [PubMed] [Google Scholar]
- 11.Johnson EH, Gant L, Hinkle YA, Gilbert D, Willie C, Hoopwood T. Do African-American men and women differ in their knowledge about AIDS, attitudes about condoms, and sexual behaviors? J Natl Med Assoc. 1992;84:49–64. [PMC free article] [PubMed] [Google Scholar]
- 12.Neal JJ, Fleming PL, Green TA, Ward JW. Trends in heterosexually acquired AIDS in the United States, 1988 through 1995. J Acquir Immune Defic Syndr Hum Retrovirol. 1997;14:465–74. doi: 10.1097/00042560-199704150-00011. [DOI] [PubMed] [Google Scholar]
- 13.Brown LK, Lourie KJ, Flanagan P, High P. HIV-related attitudes and risk behavior of young adolescent mothers. AIDS Educ Prev. 1998;10:565–73. [PubMed] [Google Scholar]
- 14.Lugoe WL, Klepp KI, Skutle A. Sexual debut and predictors of condom use among secondary school students in Arusha, Tanzania. AIDS Care. 1996;8:443–52. doi: 10.1080/09540129650125632. [DOI] [PubMed] [Google Scholar]
- 15.Kapiga SH, Lwithula GK, Shao JF, Hunter DJ. Predictors of AIDS knowledge, condom use and high risk sexual behavior among women in Dares-es-Salam, Tanzania. Int J STD AIDS. 1995;6:175–83. doi: 10.1177/095646249500600307. [DOI] [PubMed] [Google Scholar]
- 16.Bandura A, Adams NE, Beyer J. Cognitive processes mediating behavioral changes. Journal of Personality and Social Psychology. 1977;35:125–139. doi: 10.1037//0022-3514.35.3.125. [DOI] [PubMed] [Google Scholar]
- 17.Solberg OA, Laberg JC, Bjorn HJ, Eid J. Predictors of self-efficacy in a Norwegian Battalion prior to development in an international operation. Military Psychology. 2005;17:299–314. [Google Scholar]
- 18.Bandura A. Social foundations of thought and action: A social cognitive theory. Englewood Cliffs: Prentice Hall; 1986. [Google Scholar]
- 19.Ross MW, Essien EJ, Ekong E, James TM, Amos C, Ogungbade GO, et al. The impact of a situationally focused individual human immunodeficiency virus/sexually transmitted disease risk-reduction intervention on risk behavior in a 1-year cohort of Nigerian Military personnel. Military Medicine. 2006;171:1027–1034. doi: 10.7205/milmed.171.10.970. [DOI] [PubMed] [Google Scholar]
- 20.Hosmer DW, Lemeshow S. Applied Logistic Regression. 2. Hoboken, NJ: Wiley-Interscience; 2000. [Google Scholar]
- 21.Sterk CE, Theall KP, Elifson KW, Kidder P. HIV risk reduction among AA women who inject drugs: A randomized-control trial. AIDS and Behavior. 2003;7:73–86. doi: 10.1023/a:1022565524508. [DOI] [PubMed] [Google Scholar]
- 22.Rhodes F, Wolitski RJ, Thornton-Johnson S. An experimental program to reduce AIDS risk among female sex partners of injection drug users. Health and Social work. 1992;17:261–272. doi: 10.1093/hsw/17.4.261. [DOI] [PubMed] [Google Scholar]
- 23.Sterk CE, Theall KP, Elifson KW. Effectiveness of a risk reduction intervention among African American women who use Crack Cocaine. AIDS Educ and Prevention. 2003;15:15–32. doi: 10.1521/aeap.15.1.15.23843. [DOI] [PubMed] [Google Scholar]
- 24.Rasch RF, Weisen CA, MacDonald B, Wechsberg WM, Perritt R, Dennis ML. Patterns of HIV risk and alcohol use among AA crack abusers. Drug Alcohol Depend. 2000;58:259–266. doi: 10.1016/s0376-8716(99)00098-8. [DOI] [PubMed] [Google Scholar]
- 25.Corby NH, Wolitski RJ, Thornton-Johnson S, Tanner WM. AIDS knowledge, perception of risk, and behaviors among female sex partners of injection drug users. AIDS Educ Prev. 1991;3:353–66. [PubMed] [Google Scholar]