The workshop opened with two keynote addresses that provided an overview of the colliding epidemics and syndemics of infectious diseases and noncommunicable diseases (NCDs) and the potential approaches needed to address the convergence. In the first keynote address, Tolullah Oni, clinical senior research associate in the MRC Epidemiology Unit at the University of Cambridge, provided an overarching perspective on the trends of the dual global collision and burden of infectious diseases and NCDs, and she offered new ways to address the convergence. Emily Mendenhall, Provost’s Distinguished Associate Professor at Georgetown University, delivered a keynote address that elucidated the concept of syndemics and how syndemics research could provide opportunities to consider tackling the convergence of infectious diseases and NCDs in new ways.
To establish the rationale for taking an integrated approach to address the colliding epidemics of infectious diseases and NCDs, Tolullah Oni, clinical senior research associate in the MRC Epidemiology Unit at the University of Cambridge, first described how mortality and morbidity trends have shifted in recent decades, with the contribution of NCDs to both of those burdens increasing relative to infectious diseases. She followed with a discussion of the role of false dichotomies in delaying convergent action, the range of interventions that exist for convergent action, and lessons to apply
from infectious diseases epidemic responses, as well as new approaches to convergent prevention and control.
Shifting Mortality and Morbidity Trends
Oni began by presenting data from the Disease Control Priorities (3rd edition), which shows that the global crude death rate (CDR) per 100,000 population generally improved during the period between 2000 and 2015 (Jamison et al., 2018). Looking more closely at the contribution of NCDs versus infectious diseases to the global CDR reveals shifting trends that warrant convergent action, explained Oni. In order of magnitude, the top 10 contributors to the global CDR in 2015 were ischemic heart disease, stroke, lower respiratory infections, chronic obstructive pulmonary disease, lung cancers, diabetes mellitus, dementias, diarrheal diseases, tuberculosis (TB), and road injuries (Jamison et al., 2018). Only 3 of the top 10 contributors were infectious diseases, all of which had substantial decreases in their CDRs between 2000 and 2015 (Jamison et al., 2018). In contrast, NCDs represented 7 of the top 10 contributors to the global CDR, and there were considerable increases in the CDRs for dementias, diabetes mellitus, lung cancers, and ischemic heart disease during the same period (Jamison et al., 2018).
An even more granular picture emerges when the CDR data are broken down by country income status (see Figure 2-1). In low-income countries, lower respiratory infections are associated with the largest CDR per 100,000 population, although that CDR dropped substantially between 2000 and 2015. The second and third leading contributors to CDR are stroke and ischemic heart disease, both of which had increases in CDR during that period. Similar patterns were seen both in lower-middle-income and in upper-middle-income countries during the same period. Ischemic heart disease was the largest contributor to CDR in lower-middle-income countries, with an increasing rate over the 15-year period, while lower respiratory infections, TB, and diarrheal diseases all had substantial decreases in CDR. Upper-middle-income countries had CDR increases for the top two contributors, ischemic heart disease and stroke. The CDR rates observed in high-income countries were most similar to the global averages (Jamison et al., 2018). The two leading contributors—ischemic heart disease and stroke—saw substantial decreases in CDRs, while the CDR rate for the third greatest contributor, dementias, increased by a large percentage (Jamison et al., 2018).
Similar transitional patterns are evident in morbidity trends. In South Africa, for example, HIV/AIDS, diabetes mellitus, and major depressive disorder were among the top 10 causes of disability-adjusted life years (DALYs) in 2010 that were not on the top 10 list of causes at all
in 1990 (Institute for Health Metrics and Evaluation, 2010). Between 1990 and 2010, the number of DALYs caused by diarrheal diseases and lower respiratory infections decreased (Institute for Health Metrics and Evaluation, 2010). During the same period, the following NCD causes of DALYs increased substantially (Institute for Health Metrics and Evaluation, 2010):
- Drug use disorders (around 200 percent)
- Chronic kidney diseases (around 130 percent)
- Hypertensive heart disease (around 100 percent)
- Diabetes (around 100 percent)
- Anxiety disorders (around 50 percent)
- Major depressive disorders (around 40 percent)
- chronic obstructive pulmonary disease (COPD) (around 20 percent)
- Stroke (around 10 percent)
Oni explained that infectious diseases and NCDs are conditions that co-occur but also interact, driving a rise in multimorbidity. In South Africa, the National Income Dynamics Study, which captures data about health as well as socioeconomic deprivation, found double, triple, and even quadruple multimorbidities (Weimann et al., 2016). Diabetes-hypertension was the most common double morbidity, but HIV-hypertension was also significant and underconsidered in many health systems. Significant triple multimorbidities included diabetes-hypertension-HIV and TB-diabetes-hypertension. Hypertension-diabetes-TB-HIV was a significant quadruple multimorbidity. She added that multiple regression analyses indicated spatial clustering of multimorbidities. Factors such as living in the most socioeconomically deprived areas, living in urban versus rural areas, and being obese were all significantly associated with multimorbidity.
Another emerging epidemiological trend is that multimorbidities are being seen in people of increasingly younger ages. This is counter to the norm, but Oni noted that this trend should help spur the global health community into action. A study of people receiving care for either HIV, TB, diabetes, or hypertension looked at patterns of multimorbidity across different age groups in peri-urban South Africa (Oni et al., 2015). Among people aged 18 to 35 years living with HIV, almost 20 percent had HIV and hypertension comorbidity, and around 12 percent had HIV and type 2 diabetes comorbidity, compared to rates of less than 5 percent for people of the same age without HIV (Oni et al., 2015). This is corroborated by global mortality trends, she added. One-third of deaths in lower-middle-income countries occur among people under the age of 60 years, while deaths in high-income countries primarily occur in people over 60 years (Rotheram-Borus et al., 2015).
Dispelling False Dichotomies to Move Toward Convergence
Moving toward convergence will require dispelling myths and false dichotomies about the chronicity, interactions, and risk factors related to infectious diseases and NCDs, said Oni. As an example, she noted that chronicity was generally considered to be an exclusive feature of NCDs until HIV changed that picture. At the center of the global epidemic stage was an infectious disease that is actually a chronic disease, which catalyzed a shift in thinking about how to care for its comorbidity patterns. Among people receiving treatment for HIV, around 20 percent had another health condition: hypertension (77 percent), TB (24 percent), and diabetes mellitus (17 percent) (Oni et al., 2015).
Oni explained that a false dichotomy also pertains to disease–disease interactions. This is the presumption that the causal relationship between NCDs and infectious diseases only works in one direction. However,
evidence suggests that the relationship is bidirectional. A 2015 study of population-attributable risk to TB worldwide found the emergence of smoking and alcohol abuse as factors, in addition to the previously known contributors of HIV and undernourishment (Bates et al., 2015). A more recent study of the patterns of the epidemiology of HIV, TB, and diabetes found a 14 percent population-attributable risk fraction of TB attributable to diabetes and in the context of HIV coinfection (Oni et al., 2017). She added that the reverse also holds true—NCDs can be a consequence of infectious diseases. Furthermore, interactions among co-occurring infections and NCDs can occur over the life course. For instance, malnutrition, stunting, and repeated enteric infections can affect cardiovascular risk later in life (Oni and Unwin, 2015).
Another false dichotomy noted in Oni’s presentation is that NCDs and infectious diseases have distinct risk factors. In fact, shared risk factors are also a type of interaction between the two categories of conditions. Figure 2-2 illustrates a number of shared, often socioenvironmental and behavioral, risk factors in relation to both common infections and NCDs (Oni and Unwin, 2015). Considering these shared risk factors is a potential starting point for convergent action, she said. Rapid patterns of urbanization create an additional layer of complexity to shared risk factors, because many settings with emerging epidemics of NCDs coupled with high rates of mortality and morbidity are also experiencing rapid urbanization (Ezeh et al., 2017). Addressing this complexity will require identifying shared risk factors and developing convergent strategies for addressing them, she argued. For instance, urban built environments often have dense informal settlements that facilitate the transmission of infectious diseases (Patterson et al., 2017).
At the same time, unhealthy environments can impede healthy lifestyle choices and contribute to increases in NCDs and multimorbidities (Ezeh et al., 2017). Infectious and zoonotic diseases are reemerging as rapid urbanization pushes the boundaries of human settlements (Ko et al., 1999). This rapid urbanization is also creating waste and water demands that can overwhelm the inadequate urban infrastructure and contribute to the persistence of infectious diseases that should be controllable (Marsalek, 2014). Relatedly, large numbers of younger populations are drawn to urbanization, putting them at increasing levels of exposure and leading to onset of NCDs at earlier ages (Allender et al., 2011).
Spectrum of Interventions to Address the Convergence
The convergence of NCDs and infectious diseases warrants a spectrum of interventions, asserted Oni. Within this spectrum, primary prevention efforts can intervene on upstream factors and determinants that are often
shared between infectious diseases and NCDs. The spectrum extends to secondary prevention, such as screening and integrated services, through to tertiary prevention with integrated treatment and care. She emphasized the need to continue to consider health broadly as well as health care—the latter is an important but not exclusive component of health. Oni presented Table 2-1 to frame the spectrum in terms of exposures, health services needed to affect exposures, and the effect of those services on intermediate- and long-term health outcomes. She noted that traditional thinking about health services focuses on the dimensions of access with the aim of achieving long-term significantly positive effects on mortality and morbidity. However, she suggested thinking differently and broadening the concept of health services. For instance, habitation planning, transport, water and waste, and food could all be construed as health services, because the exposures that those services predetermine with respect to intermediate- and long-term effects could also affect both infectious diseases and NCDs (Weiss and McMichael, 2004; Ding et al., 2014).
|Health Services||Exposures||Intermediate Outcomes||Long-Term Outcomes|
|Water and Waste||
Physical activity opportunities
Health care episodes
Health care admissions
Acute respiratory disease
Chronic respiratory disease
Social cohesion opportunities
Physical activity opportunities
Obesity Social cohesion
Chronic respiratory disease
|Habitation (and planning)||
Social cohesion opportunities
Physical activity opportunities
Sleep and stress
Health care episodes
Health care admissions
Acute respiratory disease
Chronic respiratory disease
|Health Care (prophylaxis, treatment, palliation)||
Health care episodes
Health care admissions
NOTE: CVD = cardiovascular disease.
SOURCES: Oni presentation, June 11, 2019; Oni et al., 2019.
Oni explored how convergent action could be informed by responses to infectious disease epidemics by presenting a set of arguments made to activate political will, advocacy, research, and funding for HIV framed as a national security threat:
- The risk of exposure and threat of HIV are borderless, and HIV hits adults who are otherwise healthy, including educated populations that drive economic development.
- HIV affects people beyond marginalized populations and is overwhelming at the community scale, but even more detrimental at the macro scale.
- HIV’s effect on society projects across a long-time arc, with cross-generational risk that accumulates over time.
- High-risk behavior related to HIV spreads with urban migration and population mobility.
- HIV has a high and growing prevalence in regions and states with significant population growth and urbanization rates, such as China, India, and countries in Africa.
Oni reflected that the same characteristics apply to NCDs such as hypertension, yet global action has not been spurred for NCDs as it has been for infectious disease epidemics. She cautioned, “If we cannot galvanize action for both, we will not be able to act effectively to converge our actions.”
To explore reasons for the delay in convergent action against NCDs and infectious diseases, Oni presented some counterfactual points to consider. She drew a distinction between the understanding that a current situation is negative or abnormal as opposed to the shared perception that a situation is ordinary because it has been that way over time or it is not yet possible to predictably say or show the ways in which it is increasing. Similarly, it can be helpful when there is a direct link of cause and effect between an exposure and outcomes. If that link is more diffuse and distant, however, then there may be a lack of data to track the exposures that have a shared effect on shared outcomes. It can also be helpful when roles for interventions are clearly identified, she added, but complex adaptive systems do not necessarily allow for such clarity or delineation.
Potential New Approaches to Convergent Prevention and Control
Oni concluded with a discussion of new approaches to convergent prevention and action from a health care system perspective, which she stated are urgently needed. To that end, care for chronic conditions needs
to be integrated with care for the increasing burden of multimorbidities, rather than framing care around singular diseases per person. She argued that there is a need for this ethos to be integrated into the health care system at all levels and from all perspectives—from the effect of multiple chronic disease morbidities on patients, to health care providers’ capacity to deal with the complexities of those patients’ needs, to the biological interactions between chronic and infectious disease morbidities (Oni et al., 2014). At the policy level, she said these perspectives need to coalesce to inform integrated management and to engage patients with the health care system by incorporating disease interaction complexity with the perspectives of patients and providers (Oni et al., 2014). Oni added that public health science does not in and of itself create more equitable population health outcomes. This work can be achieved through systems for health that encompass the private sector, individuals and civil society, and policy at the global, regional, national, and subnational levels. Achieving equitable public health outcomes, she argued, will require consideration of each of those channels—for example, the influence of public health science on a population vis-à-vis the private-sector individuals and civil society.
Oni maintained that progress can be driven by fostering collaboration and building relationships to support and promote equitable population health outcomes by leveraging the beliefs, practices, and self-efficacy of individuals and civil society as well as harnessing the potential of economic and environmental drivers. In that context, she made the case for four approaches to convergent prevention and control.
First, adolescents and primary prevention warrant greater focus when considering the associations between co-occurring, preventable conditions and tools to better understand them (Oni and Unwin, 2015). She said that the scientific community has a responsibility to collaborate across sectors and work toward convergence in the intersectional science of prevention through a life-course approach at all socioecological levels, including families, communities, and broader structures.
Second, because treating comorbidities is a multidimensional science, she suggested adopting an integrated approach to capacity building by paying more heed to the workforce who will be implementing new systems and approaches. Integrated training for the new generation of researchers and health practitioners, she added, would provide them with the requisite inter- and transdisciplinary knowledge and experiences to break through silos.
Third, she suggested expanding surveillance beyond individual diseases. This could be done by building on evidence-based interventions around single outcomes to create more innovative approaches that address the need for sustained surveillance.
Fourth, she made the case for intersectoral collaboration focusing on
long-term outcomes by engaging with spatial determinants, particularly in urban settings. This could help shift the disproportionate amount of attention currently devoted to “quick wins” in public health, she said, which often neglect the delayed co-occurrence of morbidities, toward a focus on health outcomes rather than disease outcomes. Oni concluded her presentation by saying the following:
We need to be bolder in the ways we build and strengthen our systems for health, by thinking beyond just treatment and secondary prevention to research that addresses long-term and delayed co-occurrence in order to improve our health outcomes.
In her keynote address, Emily Mendenhall, Provost’s Distinguished Associate Professor at Georgetown University, explored how syndemics of infectious diseases and NCDs can inform new approaches to tackling the global burdens. She began by laying the conceptual groundwork for her discussion. Syndemic research builds on three core concepts of epidemiology: epidemics, pandemics, and endemics. An epidemic occurs when a disease has greater than expected frequency across a population. A pandemic is an epidemic that spreads across multiple populations, while an endemic is a well-established disease within a population that remains year after year. A syndemic is the dynamic relationships and synergies among a cluster of two or more epidemics and the various factors that precipitate their interaction within a population. She emphasized that the syndemic is the outcome of the interaction of the health conditions and the social and structural factors.
The concept of the syndemic originated in medical anthropology with Merrill Singer’s ethnography work on the complexity of HIV in an inner-city setting. He found that HIV could not be understood without also considering violence and substance abuse, leading him to develop the original definition of a syndemic: “When adverse social conditions, such as poverty and oppressive social relationships, stress a population, weaken its natural defenses, and expose it to a cluster of interacting diseases” (Singer, 1994). Today, syndemic research focuses on locally driven factors that drive clustering of diseases as well as the social, political, and ecological factors that drive that clustering, said Mendenhall. It is helpful to look at those factors as differential and regionally significant, she added, because the historical and political factors that drive the clustering differ from place to place. She emphasized that syndemics are not pandemics, as they are always localized
Because of the increasing attention they are receiving across disciplinary boundaries, syndemics need to be carefully defined, cautioned Mendenhall. She outlined three rules for identifying a syndemic:
- It requires two or more diseases that cluster within a population.
- Biological, psychological, or social interactions exist between these clustering diseases, with structural, social, and in some cases ecological factors precipitating the clustering.2
- A syndemic is an interaction between the disease clustering and the drivers that comes together to create a more complicated health experience than any disease or social problem would create alone.
For example, in the context of disparity-promoting factors that lead to disease clustering, two diseases can interact in adverse ways that lead to enhanced disease transmission, progression, and negative health outcomes (Singer et al., 2017).
Syndemics of Diabetes, Depression, and Infectious Disease
Syndemic research provides opportunities to consider the convergence of infectious diseases and NCDs. Mendenhall’s work on syndemics began when she explored rich epidemiological literature on the bidirectionality of depression and diabetes among women with diabetes who have lived experience of social trauma and sexual, physical, and emotional violence. In this case, the syndemic is the interaction between violence, trauma, diabetes, and depression, which is fundamental to understanding why those women have higher rates of mortality and morbidity than others with diabetes (Mendenhall et al., 2012). The socioeconomic reversal of diabetes parallels globalization, as diabetes escalates among low-income populations across the world (Hsu et al., 2012). Depression and poverty are also
1 Mendenhall mentioned the various contributions of syndemic scholarship to date—from theorizing the complexities of what syndemics are, especially relative to the social and medical conditions concurrent to HIV (Singer, 1994, 1996, 2009) to theoretical consideration of how syndemics may have emerged through historical and ecological lenses (Singer, 2009). Syndemic thinking has been translated to global health and clinical medicine (Mendenhall, 2017; Mendenhall et al., 2017; Singer et al., 2017; Tsai et al., 2017; Willen et al., 2017). The efficacy of current quantitative approaches to syndemic analyses has also been reconsidered (Stall et al., 2003; Tsai and Burns, 2015; Tsai, 2018).
2 An example of this is that climate change can drive the clustering of syndemics, such as when mosquitoes move to higher altitudes, thus increasing the risk of malaria and HIV infection.
strongly correlated (Heflin and Iceland, 2009). Although there may be dissimilar rates of acute depression among people with diabetes across income groups, people with untreated, unrecognized depression among low-income groups have much higher rates of diabetes-related morbidity and mortality (Mendenhall et al., 2012). These trends are driven by high lifetime levels of social stress, the interface with depression, the interface with chronic infections such as HIV and TB, and unreliable access to quality, affordable care (Reddy et al., 2007; Mendenhall et al., 2012). This type of syndemic is much more complicated than a straightforward comorbidity, she added. A study that examined the prevalence of type 2 diabetes, HIV/AIDS, TB, and depression in India, Kenya, South Africa, and the United States reveals that all four conditions are much more prevalent among low-income urban populations; those conditions also tended to cluster within the same populations (Mendenhall et al., 2017).
Mendenhall suggested that diabetes is actually a different condition across different contexts. This is because global narratives of diabetes are linked to different food systems and changes across the globe, but also because diabetes is syndemic when it is experienced and embodied by people in a community. She explained:
The trauma of forced relocation and fragmented social worlds cannot be removed from this story, because real biological and pathological pathways of trauma and chronic stress lead to diabetes, from epigenetics to psychophysiology.
Understanding those syndemic interactions allows for setting-specific understanding of diabetes and its treatment. The population of the Somali region of Ethiopia, for example, has different interactions of food and nutrition (and even wasting) with diabetes, which underscores the need to understand the psychophysiology of trauma, crisis, hunger, and stress that are fueling various types of diabetes (Carruth and Mendenhall, 2019). In a community in South Africa, Mendenhall observed a phenomenon that pushed her to think about syndemic interactions in a different way. Faith healers in the community had been educating people that HIV was a condition just like cancer or diabetes, so people with diabetes would hide their diagnosis because of fears that other people would think they actually had HIV (Mendenhall and Norris, 2015). The social contagion of stigma linked to the experiences of diabetes and cancer in those communities is another interaction that prevents people from seeking care in these contexts, she added.
To illustrate the complexity of syndemics, Mendenhall presented the narrative of Esther, a woman living with diabetes and HIV in Nairobi, Kenya (see Box 2-1). Esther’s story unpacks the external environment and
complexities driving these conditions, and depicts a larger epidemiological narrative that extends over generations. Diabetes and infections are also syndemic in settings with lower HIV presence, but higher rates of other infections (Mendenhall et al., 2017). As diabetes emerges within the context of crisis, displacement, and hunger, the condition will converge and interact with other common infections in those contexts, such as malaria and TB. In the Somali region of Ethiopia, for example, diabetes prevalence is not
well understood. An exploratory study found that most people diagnosed with diabetes did not know if they had type 1 or type 2 (Carruth and Mendenhall, 2019). Although most people were adherent to their medication, they were often underweight (Carruth and Mendenhall, 2019). This underscores the need to better understand the complex interactions between chronic infections, chronic malnutrition, and diabetes.
Why Syndemics Matter
Mendenhall explained that syndemics can change the way we think about disease. Syndemics inherently require consideration of the social, political, ecological, and embodied experiences of what it means to have an illness and what that means in a life. A tenet of anthropology is that a disease is an objective diagnosis, but an illness is experienced. Syndemics require the recognition that diseases rarely exist in isolation, and the identification of social, political, economic, and ecological factors are driving poor health. She noted that work is ongoing to find ways to test syndemics productively to inform upstream interventions and downstream integrated, universal health care. Syndemics also help to determine how co-occurring diseases, medications, social dynamics, and clinical barriers can actually make people sicker—powerful social dynamics underpin treatment adherence and compliance. The most effective interventions to mitigate syndemic interactions may be upstream, downstream, clinical, or community-based interventions. Syndemics also demand that social policies are recognized as key health interventions in people-centered care. For example, policies to promote desegregation, education, and school lunch programs can mitigate barriers to receiving care.
Mendenhall concluded that the concept of syndemics is framed within a transdisciplinary agenda. Syndemics shift thinking about disease and co-occurrence, putting context at the center of how disease emergence, experience, and intervention are understood. Cultivating a comprehensive view of how syndemics emerge, converge, interact, and change reveals how syndemics can affect populations differently, or affect the same population in different ways over time. Syndemics can shift how diseases are studied, moving from ethnography to epidemiology then back to ethnography, with complex biological, psychological, and social interactions assessed using quantitative analyses to better understand illnesses.
During the discussion, Marcos Espinal, director of communicable diseases and health analysis at the Pan American Health Organization, highlighted the importance of thinking about syndemics because it may help find
better ways to integrate and find synergies, but he also noted the importance of simplicity because many low- and middle-income countries do not always have all of the resources available. He stated that simple measures that require community engagement, such as measuring blood pressure or glucose levels at community centers, are critical as the first level of care. While there is value in drawing best practices and learning from efforts to integrate HIV and TB care, some of these approaches may not be taken up as the health workers may become fatigued with all of the different practices promoted to them. Espinal questioned, “How can we make sure health workers do their part and those simple measures are available in the community, because if we get too complex, then we might not be adopting those approaches.”
Oni suggested differentiating between complexity and complicatedness to avoid the risk of being paralyzed by complexity. Achieving these long-term goals will require embracing complexity and framing simplicity within that complexity, she said. Simplicity is vital to achieving change, she stated, but it does not mean being overly reductive. Over the long term, the focus should be on providing enough care, especially in low-resource settings, to catch up to the demand. She warned that relying on economic growth to propel change will not be sufficient—immediate action is needed to strengthen primary prevention and protect young people at high risk. Oversimplifying without embracing complexity—for example, being entirely focused on screening—will not “turn off the tap” and reduce that risk.
Mendenhall replied that inequality is driving extraordinary divides among health outcomes. For example, South Africa is a relatively wealthy country, but it has some health indicators of a low-income country that are driven (at least in part) by socioeconomic inequality (Omotoso and Koch, 2018). She added that caring for people and their general health conditions, rather than just specific diseases, could make a profound difference, but it requires larger conversations about training, resource allocation, and prioritization of strategies. She added that from an anthropological perspective, patients should feel that they are taken seriously and well cared for beyond just receiving medications; this can be addressed without requiring additional resources.
This page intentionally left blank.