The study, led by Bernard Afriyie Owusu and Benedict Weobong, focused on common mental disorders, a category that encompasses depression and anxiety, the two conditions that dominate the mental health burden in low- and middle-income countries. Rather than treating these conditions separately, the researchers constructed a composite outcome: a woman was classified as having a common mental disorder if she met standardized thresholds for probable depression, probable anxiety, or both. This approach reflects the clinical reality that depressive and anxious symptoms frequently overlap and co-occur, particularly in populations exposed to chronic stressors such as poverty, food insecurity, and violence within the home.

The numbers reveal a substantial and previously underappreciated burden. Probable depression was detected in 8.6 percent of the women surveyed, with a 95 percent confidence interval of 7.1 to 10.5 percent, while probable anxiety affected 10.7 percent, with a confidence interval of 8.7 to 12.9 percent. Taken together, 13.5 percent of partnered women carried a probable common mental disorder, and 5.8 percent reported symptoms of both depression and anxiety simultaneously. Behind each of these percentages stands a woman navigating daily life while carrying a psychological weight that, in many cases, has never been recognized or treated by any health service.

Violence within intimate relationships proved disturbingly common in the same population. Emotional violence was the most prevalent form, experienced by 16.1 percent of women, followed by physical violence at 12.0 percent and sexual violence at 4.0 percent. Crucially, the researchers did not simply ask whether violence had ever occurred. They stratified exposure into three categories: never, in the past, or recently. This temporal distinction matters enormously for understanding how violence affects the mind, because it allows researchers to separate the acute psychological impact of ongoing abuse from the longer-term consequences of violence experienced years earlier.

When the researchers ran weighted binary logistic regression models in STATA 18/MP, adjusting for a range of potential confounders, the association between recent violence and mental illness emerged with striking clarity. Women who had experienced any form of intimate partner violence recently had roughly double the odds of a common mental disorder compared with women who had never experienced violence, with an adjusted odds ratio of 2.05 and a confidence interval of 1.36 to 3.09. Women whose exposure to violence was in the past also showed elevated odds, at 1.40, though this estimate crossed the line of statistical significance with a confidence interval of 0.80 to 2.44. The pattern suggests that recent abuse drives much of the mental health burden, while past violence may leave a residual, if harder to measure, imprint on psychological wellbeing.

The most consequential finding, however, concerned the type of violence rather than its timing. After adjustment, emotional intimate partner violence remained independently associated with common mental disorders, carrying an adjusted odds ratio of 2.34 with a confidence interval of 1.6 to 3.4. Physical violence, by contrast, showed a weaker and statistically nonsignificant association, with an odds ratio of 1.21 and a confidence interval spanning 0.73 to 2.03, and sexual violence similarly failed to reach significance at an odds ratio of 1.67 with a confidence interval of 0.83 to 3.35. This hierarchy challenges a common intuition that physical harm is the most psychologically destructive form of abuse. Words that demean, isolate, threaten, and control, it appears, may cut deeper than blows, perhaps because emotional abuse tends to be chronic, erodes a woman’s sense of self-worth over months and years, and often operates in the absence of any external evidence that would prompt intervention by family, neighbors, or authorities.

The analysis also surfaced findings about socioeconomic factors that complicate any simple narrative linking poverty alone to mental illness. Higher household wealth was strongly protective, with women in wealthier households showing an adjusted odds ratio of 0.34, and a confidence interval of 0.13 to 0.88, for common mental disorders compared with the poorest women. Economic resources, the data suggest, buffer against psychological distress, whether by reducing the everyday stressors that accumulate into anxiety and depression or by expanding access to coping strategies and social support. Yet one counterintuitive result stood out: women who reported joint household decision-making with their partners had higher odds of common mental disorders, at an adjusted odds ratio of 1.84 with a confidence interval of 1.08 to 3.15. The authors of the study do not overinterpret this association, but it serves as a reminder that survey cross-sections can conceal complex dynamics, and that measures intended to capture women’s empowerment may sometimes correlate with circumstances, such as employment outside the home or exposure to conflicting family expectations, that carry their own psychological costs.

Perhaps the most theoretically intriguing result involves the interaction between marital status and violence. The researchers found evidence that the effect of intimate partner violence on mental health varies depending on a woman’s marital situation, providing partial empirical support for intersectionality theory, the framework holding that social identities and disadvantages do not operate independently but combine and interact to shape individual outcomes. In practical terms, this means that the mental health consequences of abuse cannot be understood by looking at violence in isolation. A widow, a woman in a polygamous union, and a woman in a first monogamous marriage may experience and internalize the same form of abuse in profoundly different ways, shaped by the social position each occupies within Mozambican society.

For policymakers and public health practitioners, the implications are direct. Mozambique, like many countries in sub-Saharan Africa, faces a severe shortage of mental health professionals, and common mental disorders routinely go undiagnosed in primary care settings. The study’s authors argue that their findings underscore the need for regular mental health support for women and for targeted campaigns against intimate partner violence, with particular emphasis on emotional abuse, which is both the most prevalent form of violence in this population and the form most strongly linked to psychological illness. Screening women who present with depression or anxiety for violence at home, and conversely offering psychological support to women identified as victims of abuse, could create two entry points into a system of care that currently reaches only a small fraction of those in need.

The research also carries a broader message for the global conversation on violence against women. International attention and funding have historically concentrated on physical and sexual violence, the forms most visible to medical systems and easiest to document in forensic settings. This study from Mozambique adds to a growing body of evidence that emotional violence, precisely because it is pervasive, normalized, and invisible, may constitute the largest mental health threat within intimate relationships. As demographic and health surveys across the region continue to collect standardized data on both violence and mental health, researchers will be able to test whether the pattern observed in Mozambique holds elsewhere. For now, the message for the roughly one in seven partnered Mozambican women living with a probable common mental disorder is sobering but clear: the violence that harms them most may be the kind no one else can see.

Subject of Research: The association between intimate partner violence and common mental disorders among partnered women in Mozambique

Article Title: Intimate partner violence and common mental disorders among women in Mozambique: Evidence from the 2022-2023 Mozambique Demographic and Health Survey

Article References: Owusu, B. A., & Weobong, B. (2026). Intimate partner violence and common mental disorders among women in Mozambique: Evidence from the 2022-2023 Mozambique Demographic and Health Survey. PLOS Mental Health, 3(10), e0000660. https://doi.org/10.1371/journal.pmen.0000660

Image Credits: AI Generated

DOI: 10.1371/journal.pmen.0000660

Keywords: intimate partner violence, common mental disorders, depression, anxiety, Mozambique, Demographic and Health Survey, emotional abuse, women's mental health, public health, intersectionality, logistic regression, sub-Saharan Africa

The research team, led by investigators at Guangzhou Medical University with collaborators at The Hong Kong Polytechnic University and The Chinese University of Hong Kong, analyzed data from 47,939 women aged 50 to 74, the age band in which routine mammography is most strongly recommended. Poor adherence was defined conservatively as a self-reported absence of mammography within the previous two years. Even after statistical adjustment for the many confounders that typically muddy observational research, 21.5 percent of respondents fell into the non-adherent category, and the gradient across adversity scores remained robust. This is not a marginal effect confined to a small, severely affected subgroup; it is a population-level signal that touches roughly one in five eligible women.

Adverse childhood experiences, commonly abbreviated as ACEs, encompass a range of early-life stressors including physical, emotional, and sexual abuse, emotional and physical neglect, household substance abuse, household mental illness, parental separation or divorce, and witnessing domestic violence. Decades of research have tied cumulative ACE exposure to depression, substance use, cardiovascular disease, and early mortality, but the specific pathway from childhood adversity to cancer screening behavior has remained comparatively underexplored. The new study addresses that gap directly, and its findings suggest that the long shadow of childhood trauma may extend all the way into the radiology suite, quietly shaping whether women show up for a test that can detect breast cancer at its most treatable stages.

Methodologically, the study went beyond the conventional tally of adversities. The researchers deployed logistic regression to quantify the association between ACE scores and screening non-adherence, but then turned to a more sophisticated analytical framework borrowed from statistical physics: the Ising model. Originally developed to describe interactions among magnetic spins, the Ising network approach treats each adversity type as a node in a system, with pairwise statistical dependencies as the connections between nodes. This allows researchers to map the architecture of adversity itself, revealing which experiences act as hubs that hold the network together and which are peripheral. The team then applied the NodeIdentifyR algorithm, or NIRA, to simulate what would happen if individual nodes were targeted for intervention, effectively running virtual experiments on the network before any real-world program is designed.

The network analysis produced findings with immediate practical implications. Household substance abuse emerged as the node with the highest betweenness centrality and expected influence, meaning it sat on the shortest paths connecting most other adversities and carried the greatest capacity to activate the network as a whole. In plain terms, growing up in a home where drugs or alcohol were abused appears to function as a gateway experience, statistically intertwined with nearly every other form of childhood adversity and potentially amplifying their downstream effects. Meanwhile, the simulation showed that reducing the impact of emotional abuse would yield the greatest projected decrease in a woman’s cumulative adversity load, making it the single most efficient target for intervention within this network.

Perhaps the most striking individual finding concerned sexual abuse, which showed the strongest independent association with screening non-adherence of any single ACE category. This detail matters because it suggests that certain forms of trauma may carry consequences that persist even when the broader adversity count is taken into account. Clinicians and screening program designers have long recognized that mammography can be an intimate, uncomfortable, and psychologically loaded procedure; for survivors of sexual violence, the experience of a mammogram may evoke memories and anxieties that make avoidance feel protective. The study’s results give that clinical intuition quantitative weight and argue for screening environments that are sensitive to trauma histories rather than blind to them.

The concept of trauma-informed care has gained traction across health systems in recent years, but this study provides one of the clearest quantitative rationales yet for applying it to cancer prevention. A trauma-informed mammography program might include routine, private inquiry about past abuse; the option to have a female technologist; clear narration of each step of the procedure; and explicit permission to pause or stop at any moment. It might also involve longer appointment slots, sensitivity training for staff, and follow-up protocols that treat missed appointments not as patient negligence but as a potential signal of unaddressed psychological burden. The authors’ conclusion points in exactly this direction: strategies addressing the sequelae of emotional abuse, in particular, may most effectively improve screening uptake among women with high adversity burdens.

The study’s scale and design lend it credibility, but its limitations deserve honest acknowledgment. The data are cross-sectional and self-reported, drawn from survey waves in 2020, 2022, and 2024, which means the analysis captures association rather than proven causation. ACE exposure was recalled retrospectively by adults, and recall bias, particularly among respondents with depression or other trauma-related conditions, could distort the measured relationships. Self-reported screening dates may also diverge from medical records. Because the analysis was secondary and based on publicly available, de-identified data, the authors note that institutional review board approval and additional informed consent were not required. None of these caveats erases the central signal, but they do mean that intervention trials, rather than further observational work, are the logical next step.

The public health stakes are considerable. Breast cancer remains one of the most common cancers among women worldwide, and mammography screening is among the best-evidenced tools for early detection, yet adherence rates persistently lag behind targets in many populations. If more than a fifth of eligible women in this large sample were non-adherent, and if childhood adversity meaningfully raises that risk, then screening programs that ignore trauma histories are systematically missing the women who need outreach most. The dose-response structure of the finding also implies a triage opportunity: a simple ACE questionnaire, administered in primary care, could flag patients who warrant more intensive reminder systems, patient navigation, or counseling before their next screening window closes.

What makes this research resonate beyond epidemiology is its reframing of a familiar problem. Missed mammograms are often attributed to cost, access, insurance status, or simple forgetfulness, and those factors certainly matter. But this study adds a developmental dimension that has been largely invisible in screening policy: the behavioral fingerprints of childhood adversity, laid down decades before a woman becomes eligible for mammography, may quietly govern her engagement with preventive medicine. By combining a classic dose-response analysis with network science and intervention simulation, the researchers have moved the conversation from whether adversity matters to which specific adversities matter most and where a targeted intervention would do the greatest good. That shift, from description to actionable targeting, is precisely what translational psychiatry promises, and it offers screening programs a concrete starting point: address the network at its hubs, beginning with household substance abuse and emotional abuse, and the entire structure of risk may begin to loosen.

Subject of Research: The association between cumulative adverse childhood experiences and non-adherence to mammography-based breast cancer screening among middle-aged women.

Article Title: Cumulative burden of adverse childhood experiences (ACEs) and non-adherence to breast cancer screening: A dose-response and network intervention analysis

Article References: Chen, F., Shao, J., Cai, J., Ho, Y.-S., Takemura, N., Yorke, J., Li, J., & Ye, Z. (2026). Cumulative burden of adverse childhood experiences (ACEs) and non-adherence to breast cancer screening: A dose-response and network intervention analysis. Translational Psychiatry. https://doi.org/10.1038/s41398-026-04523-3

Image Credits: AI Generated

DOI: 10.1038/s41398-026-04523-3

Keywords: adverse childhood experiences, breast cancer screening, mammography, dose-response, Ising model, network analysis, trauma-informed care, BRFSS, preventive care, emotional abuse, household substance abuse, health behavior

Network analysis has transformed how psychologists think about mental disorders over the past decade. Rather than treating conditions like anxiety as underlying diseases that cause visible symptoms, the network approach views each symptom — worry, restlessness, irritability, sleep disturbance — as a node in a web of mutually reinforcing connections. When connections between symptoms grow strong, activation can cascade through the network, making it harder to calm down once it has been triggered. In this framework, psychopathology is not a hidden cause but an emergent property of the network itself, a perspective championed by researchers such as Denny Borsboom and elaborated in methodological reviews that have made network psychometrics a staple of modern mental health research.

Network temperature extends this idea by borrowing a metaphor from statistical physics. Just as the temperature of a physical system reflects how energetic and disordered its particles are, the temperature of a psychological network quantifies its dynamic stability. A low-temperature network is rigid and settled: its nodes tend to remain in stable states, and disturbances die out quickly. A high-temperature network is volatile: nodes flip between states more freely, activations propagate, and the whole system is prone to sudden reorganization. The metric was originally proposed in the engineering literature as a general statistical index for measuring and managing networks, and it has recently been adapted to depression symptoms across adolescence in work published in Nature Mental Health. The new study is the first to apply it to a network spanning both childhood trauma and anxiety symptoms.

To build that network, the researchers used the Ising model, a statistical framework developed for magnetism that has become a workhorse of psychological network analysis for binary data. Each node in the model represents a dichotomous symptom or trauma indicator, and the edges represent the statistical tendency of pairs of nodes to co-occur, controlling for all other nodes. Trauma was measured with items reflecting the major domains of childhood maltreatment — physical abuse, emotional abuse, sexual abuse, physical neglect, and emotional neglect — while anxiety was assessed with the seven-item Generalized Anxiety Disorder scale, a widely validated screening instrument in Chinese adolescent populations. The resulting network joined the two clusters into a single connected system whose overall temperature could be computed and compared across developmental stages.

The team divided adolescence into three conventional phases — early, middle, and late — and estimated the network separately within each stage, repeating the analysis for males and females. Because the sample was so large, the estimates were unusually stable, an important consideration in network psychometrics, where small samples can produce unreliable edge weights. The researchers also computed centrality and bridge statistics, which identify which nodes are most influential within the network as a whole and which serve as the strongest conduits between otherwise distinct clusters of nodes.

Two findings stood out. First, physical abuse emerged as the most globally central node in the entire trauma–anxiety network, meaning it maintained the strongest and most numerous connections to other symptoms across developmental stages. Second, emotional abuse was identified as the most critical bridging node — the element that most tightly coupled the trauma cluster to the anxiety cluster. This distinction matters clinically. Central nodes are candidates for interventions aimed at reducing overall symptom load, while bridge nodes are the pathways through which problems in one domain spill over into another. If emotional abuse is the strongest bridge, then the psychological residue of verbal aggression, humiliation, and rejection in childhood may be the primary channel through which early adversity continues to feed anxious symptoms years later.

The developmental pattern was equally striking. At the population level, network temperature peaked during middle adolescence and declined thereafter. In other words, the trauma–anxiety system was most volatile — most prone to activation spreading and state changes — in the middle of the teenage years, and became more stable and settled as adolescents moved into late adolescence. This aligns with a broad body of developmental neuroscience describing adolescence, and especially the mid-teen years, as a period of heightened emotional reactivity and ongoing maturation of emotion regulation capacities, when the social-affective circuits of the brain are remodelled faster than the cognitive control systems that regulate them.

But the population-level curve concealed a profound sex difference. When the analyses were stratified, girls and boys followed divergent trajectories. Among females, network temperature continued to rise after middle adolescence, indicating that the trauma–anxiety network became progressively more unstable as girls moved into their later teens. Among males, by contrast, temperature dropped sharply after middle adolescence, suggesting a settling of the system. This divergence echoes epidemiological patterns showing that anxiety disorders become markedly more prevalent in girls during adolescence, and it suggests that the instability of the underlying symptom network may be one mechanism contributing to that widening gap.

The authors emphasize that network temperature offers something that conventional comparisons of symptom means cannot: a system-level characterization of stability. Two groups might report similar average levels of anxiety, yet differ dramatically in how tightly coupled and volatile their symptom networks are — and that difference could determine who recovers from a bad week and who spirals into a persistent disorder. By quantifying temperature across developmental stages, the study provides a way of asking not just how much distress adolescents report, but how dynamically precarious the architecture beneath that distress happens to be at each age.

Several caveats temper the conclusions. The data are cross-sectional, so each developmental stage was assessed in different individuals rather than the same individuals followed over time; cohort effects cannot be excluded, and true within-person trajectories may differ. The sample was drawn from a single country, and cultural context shapes both the experience and the reporting of trauma and anxiety. The trauma measures were retrospective self-reports, which are subject to recall biases. Even so, the scale of the sample, the novelty of the temperature metric, and the consistency of the sex-specific patterns make this a landmark demonstration. If the findings hold up in longitudinal designs, they point toward a concrete clinical implication: screening and preventive efforts for adolescents with trauma histories may need to be timed and tailored differently for boys and girls, with the middle teenage years representing a window of maximal instability — and, perhaps, of maximal opportunity for intervention before a volatile network hardens into a stable one.

Subject of Research: Developmental and sex differences in the network dynamics linking childhood trauma to adolescent anxiety, measured with network temperature

Article Title: Developmental Stage Differences in the Childhood Trauma–Adolescent Anxiety Network: A Large-Scale Cross-Sectional Application of Network Temperature

Article References: Developmental Stage Differences in the Childhood Trauma–Adolescent Anxiety Network: A Large-Scale Cross-Sectional Application of Network Temperature. (n.d.). https://doi.org/10.1007/s10578-026-02102-7

Image Credits: AI Generated

DOI: 10.1007/s10578-026-02102-7

Keywords: childhood trauma, adolescent anxiety, network analysis, network temperature, Ising model, physical abuse, emotional abuse, adolescence, sex differences, psychopathology networks, mental health, developmental psychology