Food Insufficiency is Associated with High Risk Sexual Behavior among Women in Botswana and Swaziland Presented by: Nthabiseng Phaladze, PhD and Sheri Weiser, MD, MPH Authors: Sheri Weiser, Karen Leiter, David Bangsberg, Lisa Butler, Fiona Percy, Zakhe Hlanze, Nthabiseng Phaladze, Vincent Iacopino and Michele Heisler PLoS Medicine . October 2007; 4(10); e-260 October 22, 2007 Funded by: Physicians for Human Rights and NIH
HIV and Food Insufficiency Botswana and Swaziland have highest HIV prevalence in the world Food insufficiency leading cause of morbidity and mortality in southern Africa, linked to HIV pandemic Both food insufficiency and HIV transmission risk affected by gender inequalities Growing recognition that food insufficiency may impact HIV transmission
Food Insufficiency and HIV Transmission Risk Food insufficiency postulated to increase sexual-risk taking Malnutrition weakens immune system and compromises mucosal integrity Little data on the independent effect of food insufficiency on sexual risk-taking
Aims for Population-based Study in Botswana/Swaziland Assess association between food insufficiency and risky sexual behaviors  Determine whether gender modifies associations  Evaluate extent to which associations are mediated by SES (income, education)
Methods Cross-sectional population-based study in 5 districts of Botswana and all 4 districts in Swaziland Stratified 2-stage probability design  Inclusion criteria:  18 to 49 years old fluent in Setswana/Siswati or English  resident of Botswana/Swaziland
Methods cont. Primary independent variable : Food insufficiency (inadequate food to eat over previous 12 months) Primary outcomes: Unprotected sex with a non-monogamous partner Selling or paying for sex using money or resources Intergenerational sex (>=10 years age difference) Lack of control in sexual relationships
Analysis Multivariate logistic regression, stratified by gender Standard errors were adjusted to account for effects of clustering by country* Covariates  Demographics (including income and education) HIV Knowledge (based on UNAIDS measure)**  Problem Drinking (>7 drinks/week for women, 14 drinks/week for men) **UNAIDS HIV/AIDS Prevention Indicator Survey; May 2000 *Using the Huber/White heteroscedastic estimator of the variance/covariance
Sample Conducted between  November 2004 and  May 2005 2,309 individuals randomly selected (Botswana 1433, Swaziland 876)  2,049 included, 89% response rate (Swaziland 91%, Botswana 88%)
Participant Characteristics * Denotes statistically significant differences between men and women 52%* 47%* Education >=HS 10%* 2%* Paying for sex 2%* 26%* Lack of control in sex 15%  17%  Intergenerational sex 1%* 5%* Exchanging sex 11%* 8%* Unprotected sex 30%*  17%*  Problem drinking 38%  32%  Rural residence 22%* 32%* Food insufficiency Men  (n=999) Women  (n=1050)
Adjusted Correlates of Unprotected Sex with Non-monogamous Partner* *Model also controlled for age and  living with sexual partner 1.1 (1.1-1.2) 1.7 (1.3-2.4) Food insufficiency 2.1 (1.4-3.2) 4.4 (1.7-11.0) Problem drinking 1.0 (0.8-1.2) 0.6 (0.5-0.6) Correct HIV knowledge 1.1 (0.3-3.7) 0.5 (0.4-0.5) Rural residence 0.8 (0.4-1.6) 1.1 (0.8-1.6) Married (compared with single) 0.8 (0.7-0.8) 1.2 (0.6-2.4) Income >= Median 0.9 (0.6-1.4) 0.7 (0.6-0.9) Education >= HS Men AOR (95% CI) Women AOR (95% CI)
Adjusted Correlates of Sex Exchange* *Model also controlled for age and living with sexual partner 1.1 (0.5-2.3) 1.8 (1.7-1.9) Food insufficiency 3.2 (1.9-5.5) 12.5 (8.5-18.3) Problem drinking 1.5 (0.3-6.6) 0.7 (0.6-0.8) Correct HIV knowledge 1.1 (0.6-2.2) 1.7 (1.0-3.1) Rural residence 0.2 (0.1-0.6) 0.8 (0.6-1.0) Married (compared with single) 1.9 (1.6-2.4) 0.9 (0.5-1.6) Income >= Median 1.1 (0.8-1.6) 0.7 (0.4-1.5) Education >= HS Men (Paying for sex) AOR (95% CI) Women  (Exchanging sex) AOR (95% CI)
Adjusted Correlates of Intergenerational Sex* *Model also controlled for age and living with sexual partner 1.3 (0.6-2.7) 1.5 (1.1-2.1) Food insufficiency 1.0 (0.5-2.1) 2.1 (1.6-2.8) Problem drinking 1.4 (1.0-2.1) 1.3 (0.9-1.8) Correct HIV knowledge 1.1 (0.7-1.3) 1.2 (0.8-1.7) Rural residence 0.3 (0.1-1.3) 1.1 (0.5-2.1) Married (compared with single) 1.1 (0.4-3.1) 0.9 (0.7-1.2) Income >= Median 0.8 (0.4-1.8) 0.8 (0.5-1.3) Education >= HS Men AOR (95% CI) Women  AOR (95% CI)
Adjusted Correlates of Lack of Control in Sexual Relations for Women* *Model also controlled for age and living with sexual partner 1.2 - 2.3 1.7 Food insufficiency 0.9 - 2.0 1.3 Problem drinking 1.1 - 1.7 1.4 Correct HIV knowledge 0.8 - 1.7 1.2 Rural residence 2.0 - 2.5 2.3 Married (compared with single) 0.8 - 1.1 0.9 Income >= Median 0.4 - 0.4 0.4 Education >= HS 95% CI AOR
Limitations Social desirability bias Cross-sectional design limits ability to infer causality SES measured by income and education only  Food insufficiency measure possibly insensitive
Conclusions High prevalence of food insufficiency  (22% men, 32% women) Strong and consistent relationship between food insufficiency and multiple high risk sexual behaviors among women Relationship between food insufficiency and risky sex for women not mediated by income and education
Implications  Protecting and promoting right to food may decrease vulnerability to HIV Integration of food assistance and HIV prevention programs may help reduce HIV transmission risk
Acknowledgements Affiliations Physicians for Human Rights University of California, San Francisco University of Botswana Women and Law, Swaziland University of Michigan, School of Medicine Funding sources Physicians for Human Rights NIMH Acknowledgements:   Dr. Sheila Tlou, Dr. William Wolfe, David Tuller,  Dr. Ibou Thior, Dr. Diane Havlir Dr. Vijai Dwivedi, Dr. Wayne Steward, Dr. Banu Kahn, Dr. Edward Frongillo, Dr. Steven Morin, Dr. Susan Kegeles, Dr. Willi MacFarland, David Ngele, Dr. Diana Dickinson, Choice Ginindza, Sibongile Maseko, Dr. Donald De Korte, Leonard Rubenstein.

Food Insufficiency is Associated with High Risk Sexual Behavior among Women in Botswana and Swaziland

  • 1.
    Food Insufficiency isAssociated with High Risk Sexual Behavior among Women in Botswana and Swaziland Presented by: Nthabiseng Phaladze, PhD and Sheri Weiser, MD, MPH Authors: Sheri Weiser, Karen Leiter, David Bangsberg, Lisa Butler, Fiona Percy, Zakhe Hlanze, Nthabiseng Phaladze, Vincent Iacopino and Michele Heisler PLoS Medicine . October 2007; 4(10); e-260 October 22, 2007 Funded by: Physicians for Human Rights and NIH
  • 2.
    HIV and FoodInsufficiency Botswana and Swaziland have highest HIV prevalence in the world Food insufficiency leading cause of morbidity and mortality in southern Africa, linked to HIV pandemic Both food insufficiency and HIV transmission risk affected by gender inequalities Growing recognition that food insufficiency may impact HIV transmission
  • 3.
    Food Insufficiency andHIV Transmission Risk Food insufficiency postulated to increase sexual-risk taking Malnutrition weakens immune system and compromises mucosal integrity Little data on the independent effect of food insufficiency on sexual risk-taking
  • 4.
    Aims for Population-basedStudy in Botswana/Swaziland Assess association between food insufficiency and risky sexual behaviors Determine whether gender modifies associations Evaluate extent to which associations are mediated by SES (income, education)
  • 5.
    Methods Cross-sectional population-basedstudy in 5 districts of Botswana and all 4 districts in Swaziland Stratified 2-stage probability design Inclusion criteria: 18 to 49 years old fluent in Setswana/Siswati or English resident of Botswana/Swaziland
  • 6.
    Methods cont. Primaryindependent variable : Food insufficiency (inadequate food to eat over previous 12 months) Primary outcomes: Unprotected sex with a non-monogamous partner Selling or paying for sex using money or resources Intergenerational sex (>=10 years age difference) Lack of control in sexual relationships
  • 7.
    Analysis Multivariate logisticregression, stratified by gender Standard errors were adjusted to account for effects of clustering by country* Covariates Demographics (including income and education) HIV Knowledge (based on UNAIDS measure)** Problem Drinking (>7 drinks/week for women, 14 drinks/week for men) **UNAIDS HIV/AIDS Prevention Indicator Survey; May 2000 *Using the Huber/White heteroscedastic estimator of the variance/covariance
  • 8.
    Sample Conducted between November 2004 and May 2005 2,309 individuals randomly selected (Botswana 1433, Swaziland 876) 2,049 included, 89% response rate (Swaziland 91%, Botswana 88%)
  • 9.
    Participant Characteristics *Denotes statistically significant differences between men and women 52%* 47%* Education >=HS 10%* 2%* Paying for sex 2%* 26%* Lack of control in sex 15%  17%  Intergenerational sex 1%* 5%* Exchanging sex 11%* 8%* Unprotected sex 30%*  17%*  Problem drinking 38%  32%  Rural residence 22%* 32%* Food insufficiency Men (n=999) Women (n=1050)
  • 10.
    Adjusted Correlates ofUnprotected Sex with Non-monogamous Partner* *Model also controlled for age and living with sexual partner 1.1 (1.1-1.2) 1.7 (1.3-2.4) Food insufficiency 2.1 (1.4-3.2) 4.4 (1.7-11.0) Problem drinking 1.0 (0.8-1.2) 0.6 (0.5-0.6) Correct HIV knowledge 1.1 (0.3-3.7) 0.5 (0.4-0.5) Rural residence 0.8 (0.4-1.6) 1.1 (0.8-1.6) Married (compared with single) 0.8 (0.7-0.8) 1.2 (0.6-2.4) Income >= Median 0.9 (0.6-1.4) 0.7 (0.6-0.9) Education >= HS Men AOR (95% CI) Women AOR (95% CI)
  • 11.
    Adjusted Correlates ofSex Exchange* *Model also controlled for age and living with sexual partner 1.1 (0.5-2.3) 1.8 (1.7-1.9) Food insufficiency 3.2 (1.9-5.5) 12.5 (8.5-18.3) Problem drinking 1.5 (0.3-6.6) 0.7 (0.6-0.8) Correct HIV knowledge 1.1 (0.6-2.2) 1.7 (1.0-3.1) Rural residence 0.2 (0.1-0.6) 0.8 (0.6-1.0) Married (compared with single) 1.9 (1.6-2.4) 0.9 (0.5-1.6) Income >= Median 1.1 (0.8-1.6) 0.7 (0.4-1.5) Education >= HS Men (Paying for sex) AOR (95% CI) Women (Exchanging sex) AOR (95% CI)
  • 12.
    Adjusted Correlates ofIntergenerational Sex* *Model also controlled for age and living with sexual partner 1.3 (0.6-2.7) 1.5 (1.1-2.1) Food insufficiency 1.0 (0.5-2.1) 2.1 (1.6-2.8) Problem drinking 1.4 (1.0-2.1) 1.3 (0.9-1.8) Correct HIV knowledge 1.1 (0.7-1.3) 1.2 (0.8-1.7) Rural residence 0.3 (0.1-1.3) 1.1 (0.5-2.1) Married (compared with single) 1.1 (0.4-3.1) 0.9 (0.7-1.2) Income >= Median 0.8 (0.4-1.8) 0.8 (0.5-1.3) Education >= HS Men AOR (95% CI) Women AOR (95% CI)
  • 13.
    Adjusted Correlates ofLack of Control in Sexual Relations for Women* *Model also controlled for age and living with sexual partner 1.2 - 2.3 1.7 Food insufficiency 0.9 - 2.0 1.3 Problem drinking 1.1 - 1.7 1.4 Correct HIV knowledge 0.8 - 1.7 1.2 Rural residence 2.0 - 2.5 2.3 Married (compared with single) 0.8 - 1.1 0.9 Income >= Median 0.4 - 0.4 0.4 Education >= HS 95% CI AOR
  • 14.
    Limitations Social desirabilitybias Cross-sectional design limits ability to infer causality SES measured by income and education only Food insufficiency measure possibly insensitive
  • 15.
    Conclusions High prevalenceof food insufficiency (22% men, 32% women) Strong and consistent relationship between food insufficiency and multiple high risk sexual behaviors among women Relationship between food insufficiency and risky sex for women not mediated by income and education
  • 16.
    Implications Protectingand promoting right to food may decrease vulnerability to HIV Integration of food assistance and HIV prevention programs may help reduce HIV transmission risk
  • 17.
    Acknowledgements Affiliations Physiciansfor Human Rights University of California, San Francisco University of Botswana Women and Law, Swaziland University of Michigan, School of Medicine Funding sources Physicians for Human Rights NIMH Acknowledgements: Dr. Sheila Tlou, Dr. William Wolfe, David Tuller, Dr. Ibou Thior, Dr. Diane Havlir Dr. Vijai Dwivedi, Dr. Wayne Steward, Dr. Banu Kahn, Dr. Edward Frongillo, Dr. Steven Morin, Dr. Susan Kegeles, Dr. Willi MacFarland, David Ngele, Dr. Diana Dickinson, Choice Ginindza, Sibongile Maseko, Dr. Donald De Korte, Leonard Rubenstein.