Is household food insecurity associated with social attitudes accepting of physical intimate partner violence against women in Nigeria? A population-level cross-sectional study

Discussion

We sought to assess the relationships between HFI and attitudes towards physical IPVAW in a nationally representative sample and to examine effect modification of this relationship by gender. We found that the overall prevalence of physical IPVAW acceptance was high (48%). Overall, partnered women were more likely than men to express attitudes accepting of physical IPVAW. Although puzzling, similar patterns of IPVAW acceptance have been reported in other studies conducted in Asia and Africa.15 32 33 These results are consistent with Kandiyoti’s theory of ‘patriarchal bargaining’. According to this theory, women living within patriarchal systems may feel compelled or motivated to conform to societal norms that blame wives in cases of violence as a means of coping with IPV.34 In addition, women may internalise the belief that a husband’s physical or verbal reprimand is justified and serves her interests, leading to the perception that such actions are rather legitimate responses to a wife’s perceived disobedience than acts of violence.15 34

Also, our study highlights that, overall, controlling for gender and other factors, living in households experiencing severe but not moderate FI may be influential in predicting attitudes expressing physical IPVAW acceptance. We observed distinct patterns in these associations depending on the severity of HFI and the forms of physical IPVAW acceptance. While these findings are partly consistent with our guiding hypothesis, the paucity of empirical studies focusing on FI and social norms justifying IPV precludes the extent to which we are able to compare our findings with previous literature. One study conducted in India however showed that economic debt, a proxy measure of material hardship, was associated with men’s attitudes accepting norms and promoting IPV.35 While our study did not find a significant association between household FI and attitudinal acceptance of physical IPVAW among partnered men, these disparate results collectively suggest that attitudinal acceptance of IPV may play a role in the relationship between the severity of FI and IPV as well as the role context plays in this relationship. However, since we did not perform a mediation analysis, we cannot assume these norms mediate the association between FI and actual IPV behaviours as variables on IPV experiences were not captured in the MICS dataset. Thus, this remains an open topic for future research. Another study also showed a significantly positive correlation between levels of multidimensional deprivation and attitudinal acceptance of physical IPV among men and women in 49 LMICs.34 The study, however, was not limited to individuals who were married or currently in a relationship. Also, a study by Stickley et al
36 showed that in Moscow, men who have experienced past year financial hardship had twofold higher odds of IPVAW acceptance; however, this association was not significant for women. In a conceptual framework by Hatcher et al,26 proposing the likely pathway through which FI leads to IPV perpetration by men, FI was likely to be associated with maladaptive coping strategies and consequently resulting in IPV.

Furthermore, our results indicate that gender modified the association between HFI and the likelihood of expressing attitudes accepting of physical IPVAW among food insecure individuals compared with those who were food secure. A possible explanation for this heterogeneity could be that food stress at the household level could make women perceive increased IPV vulnerability, especially within patriarchal societies characterised by unequal gender dynamics. In such contexts, women often lack autonomy and control over household resources, which could increase their susceptibility to IPV. Gendered power dynamics further reinforce traditional roles and expectations, contributing to a higher likelihood of physical IPV acceptance among food insecure women compared with men.

While our findings do not provide conclusive evidence for or against the existence of a causal relationship between HFI and physical IPVAW attitudes, they nevertheless support the need for this association to be examined in longitudinal studies, taking into consideration the intersections of other aspects of the TPB, such as subjective norms and perceived behavioural control, that could offer new insight into the mechanistic pathways linking HFI to IPVAW. In addition, as evidence suggests survivors of IPVAW are more likely to experience FI relative to the general population,37 38 future studies employing qualitative approaches can provide further explanations regarding the possibility of a bidirectional relationship between food-related hardship and social norms surrounding IPV and the specific forms. Furthermore, research should explore whether and to what extent the conclusions drawn from this study hold for attitudes towards other forms of IPVAW.

Several limitations preclude the validity of our study’s findings. As IPVAW can take many forms, our findings may not be generalisable to other forms of attitudes toward IPVAW (eg, sexual, psychological). In addition, the measures of attitudes towards physical IPV used in this study are non-exhaustive. Therefore, we are unable to establish whether the conclusions from our study would hold in the context of other scenarios, for example, ‘if wife hits him first’, ‘to discipline or keep her in line’ or ‘anytime he wants’. In addition, as this study focused on IPVAW attitudes, our findings may not hold for IPVAW intentions and actual behaviours considering that HFI may also act through other pathways as specified in the TPB. Second, an important factor to consider is that our estimates could have been severely biased due to the sensitive nature of both HFI and IPVAW, which were based on self-reports. Hence, the potential for recall bias in reporting HFI and social desirability bias in reporting IPV, which could impact our findings. Last, it is important to acknowledge that HFI was measured at the household level. However, individual experiences of food deprivation may vary even within households.39 Therefore, the generalisability of our findings to individual-level food deprivation may be limited and remains an open question.

Despite these limitations, our study has several strengths. Given the intersection of this study across several goals of the sustainable development agenda (SDG 2, 3 and 5), our findings offer comprehensive and valuable insights into the intersections between HFI, IPV and broader social and development objectives that can inform policies and interventions targeting multiple SDGs simultaneously. By using recent nationally representative data, our study provides up-to-date evidence on the prevalence of physical IPV attitudes in Nigeria and as well identifies specific subgroups within the population that are most accepting of physical IPVAW, thereby highlighting the patterns and extent to which these norms remain pervasive. Finally, our study provides reliable and robust estimates of HFI, at least to the extent to which data in the MICS are valid, based on ascertainment using a widely recognised and well-validated measure. This approach allows for comparability of our findings across various contexts and studies. Additionally, our research extends beyond the conventional approach of dichotomizing FI, which may oversimplify and mask its effect.