Transcript Gender-based violence and HIV: What don’t we know?

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
70%
65.2%
60%
50.5%
50%
43.7%
37.4%
40%
33.1%
28.5%
30%
25.2%
22.3%
20%
10%
0%
Psychological
abuse
Physical IPV
Rape
Economic abuse
Perpetration
Experience
• 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 aOR1.54 (95%CI 1.19, 1.99)
– Women who have reported less relationship power are more
likely to have HIV (aOR1.56 (95%CI1.15, 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
1.00
some 1.51 (0.65, 3.54)
often 2.13 (1.04, 4.37)
Sexual abuse: none
0.34
0.04
1.00
some 1.32 (0.88, 2.00)
often 1.66 (1.04, 2.63)
Emotional abuse: none
p value
0.18
0.03
1.00
some 1.70 (1.12, 2.57)
often 1.96 (1.25, 3.06)
0.01
0.003
Men, violence & HIV
Association between age, IPV perpetration
and HIV from logistic regresssion model
OR
Age<25 & no IPV 1.00
Age<25 & P IPV 2.08
Age>25 & no IPV 8.29
Age >25 & P IPV 10.03
95% CI
p value
1.07 4.06 0.031
5.03 13.65 <0.0001
5.74 17.52 <0.0001
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)
% sero-converting
very violent &
risky
pretty violent
moderate
3.48
2.78
2.55
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)