Intimate Partner Violence and Body Weight through Regions:
The Case of Domestic Violence in South African Women
Research Question
How does body weight affect Intimate Partner Violence (IPV) reporting throughout the regions of South Africa?
Study Design
The data used in this analysis is IPUMS DHS Global Health data that is compiled from the South Africa Demographic and Health Survey 2016. This sample looks at South African Women ages 15 to 49.
Background
South Africa is the most southern country in Africa with grasslands, deserts, forest, mountainous valleys, and beaches.
The population of South Africa is categorized as follows:
- 2018: Urban - 66.4% // Rural - 33.6%
- 2019: Over half of the population is younger than 30 years old.
- 2019: South Africa has a similar birth rate to the world average but has a death rate that is double the world average.
The ethnic composition of South Africa falls into four categories -- Black, White, Khoekhoe/San (Colored), and Asian according to South Africa's Population Registration Act. These classifications were arbitrary, but looked at family background and culture. The origins of this classification is as follows:
- Black: Bantu-speaking Africans and their descendants make up this ethnic group.
- White: Most South Africans that descended from European settlers create this category.
- Khoekhoe/San: The Khoekhoe and San peoples came to South Africa by the Dutch as slaves from Madagascar, Malaysia, Indonesia, and more. The descendants were categorized as this ethnic group. Some groups still continue to identify with "Colored" while others reject the label. Note that the original DHS data uses the term "Colored" but it has been changed to Khoekhoe/San on graphs and analysis.
- Asian: South Africans who came from Indian descent as slaves by the British were categorized under "Asian" to create the smallest group. Other communities include ethnic Asians as well.
Perceived Weight by Regions
Note that the perception of body weight is a perceived status, not a calculation of BMI or actual weight.
Women's Body Image in the Regions of South Africa
Looking overall at South Africa in this dataset from 2016, 7.42% of the women in this sample thought they were underweight. To make up the rest of the sample, 78.63% of the women perceived themselves as normal weight and 13.95% perceived themselves as overweight.
The highest percentage of women perceiving themselves as underweight is in the Eastern Cape at 14.54%. Western Cape includes the least amount of underweight perceptions with 4.08% of women.
Kwazulu-Natal has the highest amount of women perceiving themselves as normal weight at 83.77%, and Western Cape has the lowest with 69.53%.
As for the overweight perception, Western Cape has the highest amount of women with 26.53% of women, and the lowest is Kwazulu-Natal a 9.42%.
** The overweight category contains both overweight and obese perceptions.
Scroll and click for more information on South African regions.
1 | Eastern Cape
1 | Eastern Cape. Click to expand.
Underweight: 14.54%
2 | Mpumalanga
2 | Mpumalanga. Click to expand.
Underweight: 9.23%
3 | Free State
3 | Free State. Click to expand.
Underweight: 7.36%
4 | Northern Cape
4 | Northern Cape. Click to expand.
Underweight: 6.98%
5 | Kwazulu-Natal
5 | Kwazulu-Natal. Click to expand.
Underweight: 6.81%
6 | Limpopo
6 | Limpopo. Click to expand.
Underweight: 5.82%
7 | North West
7 | North West. Click to expand.
Underweight: 5.56%
8 | Gauteng
8 | Gauteng. Click to expand.
Underweight: 4.29%
9 | Western Cape
9 | Western Cape. Click to expand.
Underweight: 4.08%
1 | Eastern Cape
Underweight: 14.54%
Normal weight: 70.12%
Overweight: 15.33%
| Providential Capital: Bhisho
| Ethnicity: Blacks make up about 3/4ths of Eastern Cape's population.
|Status: Urban
2 | Mpumalanga
Underweight: 9.23%
Normal weight: 75.00%
Overweight: 15.77%
| Providential Capital: Nelspruit
| Ethnicity: Blacks, mostly Nguni, make up 9/10ths of Mpumalanga's population.
|Status: Urban
3 | Free State
Underweight: 7.36%
Normal weight: 80.29%
Overweight: 12.35%
| Providential Capital: Bloemfontein
| Ethnicity: Blacks make up 3/5ths of Free State's population.
|Status: Rural
4 | Northern Cape
Underweight: 6.98%
Normal weight: 82.96%
Overweight: 10.06%
| Providential Capital: Kimberley
| Ethnicity: Around half of Northern Cape's population is Khoekhoe/Sans.
|Status: Rural
5 | Kwazulu-Natal
Underweight: 6.81%
Normal weight: 83.77%
Overweight: 9.42%
| Providential Capital: Pietermaritzburg
| Ethnicity: Blacks descended from the Zulu peoples make up 4/5ths of Kwazulu-Natal's population. Asian/Indians make up about 1/10th of the population.
|Status: Rural
6 | Limpopo
Underweight: 5.82%
Normal weight: 79.45%
Overweight: 14.73%
| Providential Capital: Polokwane
| Ethnicity: Limpopo's population is mostly Blacks with a minority of Whites.
|Status: Rural
7 | North West
Underweight: 5.56%
Normal weight: 81.82%
Overweight: 12.63%
| Providential Capital: Mafikeng
| Ethnicity: Blacks make up 9/10ths of North West's population.
|Status: Rural
8 | Gauteng
Underweight: 4.29%
Normal Weight: 81.86%
Overweight: 13.84%
| Providential Capital: Johannesburg (South Africa's largest metropolitan city)
| Ethnicity: Blacks make up 3/4ths of Gauteng's population.
|Status: Urban
9 | Western Cape
Underweight: 4.08%
Normal weight: 69.38%
Overweight: 26.53%
| Providential Capital: Cape Town
| Ethnicity: Around half of Western Cape's population is Khoekhoe/Sans.
|Status: Urban
Experiencing IPV
Through this analysis, IPV was broken down into two categories of "emotional violence" and "physical violence" to determine differentiating factors relating to overall domestic violence.
Emotional IPV
Here are the components used to look at emotional IPV:
- Frequency of humiliation.
- Frequency of insults or making you feel bad.
- Frequency of threats of harm in last 12-months.
- Frequency of refusal to give enough money for household expenses.
- Frequency afraid of husband.
About 2 out of 3 women perceiving themselves as underweight experienced emotional IPV.
Around 6% of the normal weight women experienced emotional IPV.
Almost 3 out of 4 overweight women have experienced emotional IPV.
Here are the components used to look at physical IPV:
- Frequency of spouse's threats or attacks with gun or knife.
- Frequency spouse ever pushed, shook, or threw something.
- Frequency of husband's attempts to strangle or burn.
- Frequency of being kicked or dragged.
About 15% of underweight women have experienced physical IPV.
About 32% of normal weight women have experienced physical IPV.
Almost 7% of overweight women have experienced physical IPV.
Discussion
Logistic Regression Model
As can be seen in this Logistic Regression Model, women who perceive themselves as underweight are more likely to report IPV. This effect is persistent for physical violence, but the effect goes away for emotional violence when taking geography into account.
This shows that there is a robust effect for underweight women. These women seem to be a vulnerable population as there is a significant trend between this body weight category and reporting IPV.
One potential explanation for this could be a connection to malnutrition in the underweight women. Malnutrition, which directly relates to underweight populations, can be linked to wealth status. There was not an option to control for wealth within the survey data used, but having electric connection can be a signal for wealth. Knowing this, having electric connection in the household seems to correlate to reporting less physical and emotional violence.
When looking into the geography, reporting IPV is more pronounced in some regions over others but most are close to significant. For emotional IPV, including the geographic variables makes the effect on underweight women pass the significance mark of 0.05 but still stays relatively significant. The difference of why certain regions are significant is worthy of future further dissection.
As for the normal weight and overweight categories, these do not seem to have a large impact on reporting IPV. This finding adds to the story of what impact the perception of weight may have on reporting IPV.
Another note to consider is that the sample for ethnicity, specifically the Khoekhoe/San peoples and the Asian/Indian peoples, have very low sample sizes that could have an impact on the results. When comparing ethnicity to weight perception, very few women fit the overweight category within the different ethnicities.
References
Anon. n.d. “South Africa | History, Capital, Flag, Map, Population, & Facts.” Encyclopedia Britannica. Retrieved March 16, 2021 (https://www.britannica.com/place/South-Africa).
Ben-Tovim, David I., and M. Kay Walker. 1994. “The Influence of Age and Weight on Women’s Body Attitudes as Measured by the Body Attitudes Questionnaire (BAQ).” Journal of Psychosomatic Research 38(5):477–81. doi: 10.1016/0022-3999(94)90109-0.
Ellsberg, Mary, Lori Heise, Rodolfo Peña, Sonia Agurto, and Anna Winkvist. 2001. “Researching Domestic Violence Against Women: Methodological and Ethical Considerations.” Studies in Family Planning 32(1):1–16. doi: https://doi.org/10.1111/j.1728-4465.2001.00001.x.
Fitzgibbon, Marian L., Lisa R. Blackman, and Mary E. Avellone. 2000. “The Relationship Between Body Image Discrepancy and Body Mass Index Across Ethnic Groups.” Obesity Research 8(8):582–89. doi: https://doi.org/10.1038/oby.2000.75.
Gregory, Cria O., Heidi M. Blanck, Cathleen Gillespie, L. Michele Maynard, and Mary K. Serdula. 2008. “Health Perceptions and Demographic Characteristics Associated With Underassessment of Body Weight.” Obesity 16(5):979–86. doi: 10.1038/oby.2008.22.
Lackland, Daniel T., Trevor J. Orchard, Julian E. Keil, Donald E. Saunders, Frances C. Wheeler, Lucile L. Adams-Campbell, Robert H. McDONALD, and Rebecca G. Knapp. 1992. “Are Race Differences in the Prevalence of Hypertension Explained by Body Mass and Fat Distribution? A Survey in a Biracial Population.” International Journal of Epidemiology 21(2):236–45. doi: 10.1093/ije/21.2.236.
Mciza, Zandile, Julia H. Goedecke, Nelia P. Steyn, Karen Charlton, Thandi Puoane, Shelly Meltzer, Naomi S. Levitt, and Estelle V. Lambert. 2005. “Development and Validation of Instruments Measuring Body Image and Body Weight Dissatisfaction in South African Mothers and Their Daughters.” Public Health Nutrition 8(5):509–19. doi: 10.1079/PHN2005814.
Pearl, Rebecca L., and Rebecca M. Puhl. 2014. “Measuring Internalized Weight Attitudes across Body Weight Categories: Validation of the Modified Weight Bias Internalization Scale.” Body Image 11(1):89–92. doi: 10.1016/j.bodyim.2013.09.005.
Puoane, Thandi, Krisela Steyn, Debbie Bradshaw, Ria Laubscher, Jean Fourie, Vicki Lambert, and Nolwazi Mbananga. 2002. “Obesity in South Africa: The South African Demographic and Health Survey.” Obesity Research 10(10):1038–48. doi: https://doi.org/10.1038/oby.2002.141.
Tiggemann, Marika, and Esther D. Rothblum. 2016. “Gender Differences in Internal Beliefs About Weight and Negative Attitudes Towards Self and Others:” Psychology of Women Quarterly.
Wingood, Gina M., Ralph J. DiClemente, Kathy Harrington, and Susan L. Davies. 2002. “Body Image and African American Females’ Sexual Health.” Journal of Women’s Health & Gender-Based Medicine 11(5):433–39. doi: 10.1089/15246090260137608.
Wojcicki, Janet M. 2014. “The Double Burden Household in Sub-Saharan Africa: Maternal Overweight and Obesity and Childhood Undernutrition from the Year 2000: Results from World Health Organization Data (WHO) and Demographic Health Surveys (DHS).” BMC Public Health 14(1):1124. doi: 10.1186/1471-2458-14-1124.
Data Source For South Africa 2016:
National Department of Health (NDoH) [South Africa], Statistics South Africa (Stats SA), South African Medical Research Council (SAMRC), and ICF. South Africa Demographic and Health Survey 2016 [Dataset]. Data Extract from ZAIR71.SAV, ZAHR71.SAV, ZAKR71.SAV, ZABR32.SAV, ZAMR71.SAV, ZAPR71.SAV, and ZAAH71.SAV. IPUMS Demographic and Health Surveys (IPUMS DHS), version 7, IPUMS and ICF [Distributors]. Accessed from http://idhsdata.org on 3/27/2021.
