Linking Genderbased violence and HIV Nwabisa Jama Shai

Linking Gender-based violence and HIV Nwabisa Jama Shai Senior Researcher, Gender & Health Research Unit, Medical Research Council, Pretoria

What does gender-based violence mostly encompass in SA? Intimate partner violence: emotional, physical, sexual Sexual abuse of children Sexual violence against adults

Prevalence of gender-based violence in South Africa • Rape of women – Victimisation: 25% of women (18 -49 yrs) in Gauteng Province have been raped – Perpetration: 28 - 37% of men (18 -49 yrs) in populationbased research • Physical intimate partner violence: – Victimisation: in lifetime, disclosed by 33% of women in Gauteng Province, and in last 12 months, by 13% – Perpetration: in lifetime disclosed by 43 -51% of men, and in last 12 months by 10% • Sexual abuse of children: 39. 1% of young women from the rural E Cape has experienced sexual abuse in childhood (contact and non-contact)

Ever experienced or perpetrated GBV, adults in Gauteng, 2008

• What do we know about HIV and IPV globally from cross-sectional studies?

Gender-based violence and HIV in women • Rwanda: women in stable relationships – Phys. /sexual IPV assoc. with HIV+ (van der Straten, 1998) • Tanzania: – n=245 VCT setting, IPV assoc. with HIV+ in women <30 yrs (but not older women) (Maman et al 2002) • South Africa: – Women who have experienced > 1 episode of physical/sexual IPV a. OR 1. 54 (95%CI 1. 19, 1. 99) – Women who have reported less relationship power are more likely to have HIV (a. OR 1. 56 (95%CI 1. 15, 2. 11) (Dunkle et al 2004) • India: – n=20, 425 DHS (HIV prev. 0. 2% women) Women partners of HIV+ and violent men were 7 x more likely to have HIV than women partners of HIV+ and non-violent men (Decker et al 2009)

Multivariable poisson models of relative incidence of HIV exposure to IPV & gender inequity in a relationship (Jewkes et al, 2010) Adjusted IRR 95%CI P valve Relationship power scale: mid/high equity 1. 00 low equity 1. 51 1. 05 2. 17 0. 027 1. 04 2. 21 0. 032 Physical or sexual IPV none or 1 1. 00 >1 episode 1. 51

Poisson model, factors associated with incident HIV infections in women (n=1027) adjusted for age, treatment, stratum (partner concurrency alone was tested but is not significant) IRR 95%CI P value >1 episode of physical/sexual IPV 1. 55 1. 06, 2. 26 0. 024 Lowest relationship equity 1. 46 1. 01, 2. 10 0. 043 HSV-2 2. 29 1. 55, 3. 39 <0. 0001 Transactional sex with a casual partner during follow up 2. 06 1. 22, 3. 48 0. 007 Correct condom use at last sex before HIV result 0. 57 0. 40, 0. 82 0. 002

Incident HIV and child abuse: Stepping Stones women (Jewkes et al Child Abuse & Neglect, 2010) IRR (95%CI) Physical punishment: none Sexual abuse: none Emotional abuse: none p value 1. 00 some 1. 51 (0. 65, 3. 54) 0. 34 often 2. 13 (1. 04, 4. 37) 0. 04 1. 00 some 1. 32 (0. 88, 2. 00) 0. 18 often 1. 66 (1. 04, 2. 63) 0. 03 1. 00 some 1. 70 (1. 12, 2. 57) 0. 01 often 1. 96 (1. 25, 3. 06) 0. 003

Men, violence & HIV

Frequencies of sexual risk taking behaviours in EC/KZN men who have perpetrated >1 episode of physical IPV and those who have not Physical IPV No physical IPV p value 20+ partners ever 51. 5% 26. 0% 0. 0000 Any transactional sex 81. 0% 59. 7% 0. 0000 Sex with a prostitute 31. 6% 14. 6% 0. 0000 High levels of alcohol in past year 39. 3% 19. 2% 0. 0000 Rape of woman 49. 6% 18. 8% 0. 0000 Rape of a man 6. 6% 1. 1% 0. 0000 Consistent condom use in past year 30. 7% 41. 0% 0. 0002

Men, masculinity and HIV • Observed clustering of men’s violent and antisocial practices • In Stepping Stones study the following variables cluster into 3 groups: – Alcohol abuse, any drug use, – Emotional, physical and sexual abuse, – Gang membership, non-partner rape – Transactional sex, having 8+ life time partners • 3 groups: very violent & risky men, pretty violent & risky, and more moderate men

Relationship of class to HIV new infections (over 2 years) (latent class analysis)

Where is the evidence residing? • A very substantial body of evidence linking both gendered behaviour and GBV exposure to elevated HIV prevalence/incidence and all the supporting qualitative research • But most is from cross-sectional data, some of which is not very comprehensive… • There is a need for more longitudinal research • Current knowledge is mainly drawn from one longitudinal study – the Stepping Stones Study dataset • The data are all confirmed by cross-sectional research from Sub-Saharan Africa and India…

Key points of entry for prevention • Building gender equity: at all levels – Critically changing constructions of masculinity and acquiescent femininity • Reducing childhood exposure to GBV and sexual, physical and emotional abuse at home • Improve relationship skills: communication and conflict • Reduce substance abuse, improve access to care for mental health problems • Enhance women’s economic independence • These need to happen together

In HIV prevention terms: • Gender needs to be taken very seriously • Doing so involves addressing the context of sex and not just isolated acts or even just gender-based violence. It must include: – Relationships in general: communication, violence, gender inequity – Selfhood: think about who we are as people, how we act, how we relate to others and how we want to be seen by others – Sex and gender identities: understand sex in the context of what it means to be a man or a woman • Where ideas about gender place men and women at risk we need to target these in our prevention strategies • Zero infections requires a concerted effort to prevent GBV and to recognise it and manage the consequences in women on treatment (and reciprocal problems in men)

Conclusions • Addressing the GBV nexus is critical for achieving zero new HIV infections • There is strong evidence linking violence and gender inequity in relationships to HIV risk • Sexual practices need to be seen as flowing from gender identities, and this provides a frame for understanding why men and women behave in the way that they do (thus masculinities and femininities) • It enables reflection on the emotional and material context within which sexual behaviours are enacted, in particular the broader struggles, aspirations, desires and needs that motivate men and women’s behaviour • It follows that only when we understand this, will we be able to change sexual behaviours and thereby reduce the risk of HIV infection and improve uptake and adherence to care • Understanding individual epidemics is critical for tailored prevention • Interventions need to be theory-based at different level: – Level 1: of risk factors or drivers of the problem – Level 2: of what we seek to change (e. g. , masculinities) – Level 3: of behaviour change • What drives the behaviour • What enables change – Level 4: of how to secure change (methods or approaches – their strengths and limitations)