|Home | About | Journals | Submit | Contact Us | Français|
In August 2016, the National Heart, Lung, and Blood Institute convened a scientific forum with 18 investigators active in the field of population health disparities. The purpose was to discuss heart, lung, and sleep disorders in children and adolescents of American Indian/Alaska Native (AI/AN) heritage. All participants acknowledged the paucity of age‐ and population‐specific data on these disorders and called for new research to remedy the gap.
As participants in that forum and seasoned investigators in the field, we summarize the current state of research on cardiovascular disease (CVD) in AI/AN youth from our unique perspective. We review the available evidence on environmental exposures, diet, physical activity, and early‐onset obesity in AI/ANs, with a focus on data from individuals younger than 21 years as available, against the background of a rapidly increasing prevalence of CVD and its risk factors. It should be mentioned that the available evidence is sparse, with much of what is known about CVD in Native communities coming from the SHS (Strong Heart Study), a longitudinal, population‐based study of the epidemiology of CVD and its risk factors in adult AIs from communities in Oklahoma, Arizona, North Dakota, and South Dakota. From 1989 to 1991, SHS recruited a cohort of adults aged 45 to 75 years and followed them for several waves of data collection.1 An ancillary project, the SHFS (Strong Heart Family Study), recruited participants aged 14 years and older who belonged to the families of SHS cohort members.2 Both SHS and SHFS collected detailed information on cardiovascular events, risk factors for CVD (including genetic and environmental factors), and markers of preclinical CVD. Our review is intended to elucidate viable avenues for future research in this neglected area.
CVD was once considered rare in AIs, but lifestyle and dietary changes have contributed to a substantial increase in recent decades.3, 4 As would be expected, the population prevalence of established CVD risk factors, including obesity and diabetes mellitus, has also increased relative to the US general population.5 Despite advances in raising community awareness of CVD risk factors among AIs, the prevalence of dyslipidemia, smoking, and high blood pressure is also rising.6, 7 CVD is more likely to be fatal in AIs relative to other races, with mortality rates 20% higher than in the general population,3 while age at death because of CVD is significantly lower among Native people. Fully 36% of AIs with CVD die before age 65, versus 15% of non‐Hispanic whites.8 This premature morbidity and mortality among AI/AN adults likely reflects the high prevalence of CVD risk factors in Native youth.
Compared with other races, AI youth are disproportionally affected by obesity, which tends to begin earlier in life, as shown by data on weight from birth to preschool age.9 One longitudinal cohort study in Arizona that followed 4857 AI children (median age 11.3 years on enrollment) found that childhood obesity was associated with subsequent hypertension and insulin resistance. In addition, obesity and glucose intolerance were strongly associated with premature death from endogenous causes.10 Furthermore, a large national study found that the incidence rate of type 2 diabetes mellitus in AI/AN youth (≥19 years) rose from 22.6/100 000 in 2003 to 46.5/100 000 in 2012.11
In the SHFS cohort, 24.9% of AI adolescents met criteria for the metabolic syndrome (defined as obesity with associated hypertension, dyslipidemia, and insulin resistance).12 Metabolic syndrome is therefore substantially more prevalent in this sample than in populations of other races and ethnicities, with the National Health and Nutrition Examination Survey offering the following prevalence estimates: Mexican Americans (12.9%), non‐Hispanic whites (10.9%), and non‐Hispanic blacks (2.5%).13 Among SHFS youth with metabolic syndrome, the most prevalent risk factors were larger waist circumference, a surrogate of obesity (54.3%); low levels of high‐density lipoprotein (46.4%); hypertension (30.3%); high levels of triglycerides (27.8%); and high fasting glucose (2.5%).
SHFS results also demonstrated an association between CVD risk factors and subclinical CVD in AI youth. Similar findings have been reported for youth in the general US population.14 Obese SHFS participants displayed prevalent left ventricular hypertrophy with abnormalities in left ventricular function.15 Despite the youth of participants with hypertension, they demonstrated premature left ventricular hypertrophy with diastolic abnormalities, as well as increased arterial stiffness and elevated peripheral vascular resistance.16 Compared with participants who did not meet criteria for metabolic syndrome, those who did had a higher prevalence of cardiac abnormalities on echocardiogram, including left ventricular hypertrophy, left atrial dilation, and reduced left ventricular systolic and diastolic function.12 In the same subset of participants, prediabetes mellitus and diabetes mellitus in adolescence were associated with an untoward cardiac phenotype with left ventricular hypertrophy and left ventricular systolic and diastolic abnormalities.17 All these subclinical CVD markers are independent risk factors for adult CVD and CVD‐related mortality.18, 19, 20
Although modifiable behavioral risk factors account for a large proportion of incident CVD, increasing evidence suggests that environmental factors are also pertinent for populations around the world.21, 22, 23, 24 Such factors include the built environment, local opportunities for physical activity, access to retail outlets for healthy foods, and sociodemographic factors; however, these elements are difficult to study in rural populations. For example, quantifying the built environment relies on accurate geocoding, which is not available for AI reservations. Future study might be useful focused on neighborhood socioeconomic status in these communities.
Evidence indicates that factors such as air pollution and metal exposures also contribute to CVD, and may disproportionally affect rural populations.21, 22 For example, arsenic and cadmium lead to endothelial dysfunction, apoptosis, and vascular injury which may, in turn, exacerbate atherosclerosis.25, 26 Compared with the US all‐races population, AI communities are disproportionately exposed to several metals, including arsenic, cadmium, uranium, and tungsten.27, 28, 29, 30 For instance, SHS and SHFS participants had higher urinary concentrations of these metals, including a 30% higher concentration of cadmium and 2.4 times the amount of urinary arsenic, than did similar‐age participants in other large studies, including MESA (Multi‐Ethnic Study of Atherosclerosis), a prospective cohort study of whites, blacks, Hispanics, and Chinese Americans from 6 urban areas,28 and National Health and Nutrition Examination Survey. Other studies with AIs have also shown relatively high levels of arsenic and uranium exposure.29, 30
With regard to arsenic, uranium, and tungsten, elevated exposure might reflect groundwater contamination, which is more common in rural than in urban areas. Contamination is more likely in small community water systems in the Midwest and the Southwest and in private wells, which are currently unregulated. With regard to cadmium, both cigarette smoking and diet have been implicated as sources of exposure in the general population.31 However, elevated cadmium concentrations were found even among never‐smokers in the SHS cohort, suggesting that factors other than smoking are involved. Diet seems a more likely source of risk, as a recent study with SHFS participants found that higher urinary concentrations of cadmium were associated with higher intake of processed meats and lower intake of dairy products, fruits, and fruit juice.32
In the SHS cohort, after adjustment for sociodemographic and traditional cardiovascular risk factors, exposure to arsenic and cadmium was associated with higher incidence of cardiovascular events (including coronary heart disease and stroke) over almost 20 years of follow‐up.25, 26 For arsenic alone, the associations with CVD incidence and mortality were stronger in the subgroup of participants younger than 55 years. In the overall SHS cohort, both arsenic and cadmium were also related to higher incidence of peripheral arterial disease, while for arsenic alone, the same condition was associated with an ankle–brachial index higher than 1.4.33, 34 Similar analyses are lacking for cadmium, for which associations with carotid atherosclerosis have been reported in other populations.35, 36 Overall, the available evidence suggests that arsenic and cadmium heighten the risk of CVD in AI communities. They might also contribute to a disproportionate burden of CVD compared with other populations, including an excess of subclinical CVD in childhood and adolescence. Unfortunately, no research to date has addressed metals exposure among AI adolescents and young adults.
The evidence for disproportionate exposure to metals highlights the need for effective prevention strategies and supports ongoing efforts to install new water systems in affected AI communities. For instance, the Mni Wiconi Project has brought safe water from the Missouri River to the Oglala Sioux community in Pine Ridge.37 However, many AIs live in remote areas with limited access to community water systems and rely on unregulated private wells for drinking water. To evaluate prevention strategies for families served by such wells, an ongoing multilevel participatory intervention is testing the efficacy of a point‐of‐use drinking water filter in conjunction with an educational intervention in SHS communities in North and South Dakota. Known as the Strong Heart Water Study for Private Wells, its goal is to reduce arsenic exposure. Other strategies will also be needed, given the ubiquity of environmental exposures and the difficulty of mitigating them. One potential avenue is dietary interventions, given the evidence for associations between environmental exposures and food intake.32
In addition to metals, air pollution is an environmental exposure of major interest in AI communities. Although atmospheric pollution from vehicular traffic and industrial sources is typically lower in rural communities than in cities and suburbs, pollution from other sources might still have deleterious effects. Agriculture, livestock production, forest fires, domestic heating with coal or wood, and wind‐blown dust can all be important pollutants in rural environments. Moreover, levels of indoor air pollution can be higher than outdoor levels because of contributions from heating, cooking, and smoking tobacco.38 Solid fuel is the principal heating source for at least 2.5 million households in the United States, primarily in rural and low‐income areas, leading to high levels of indoor air pollution.39 In addition, in many “shale play” regions, unconventional gas development (often called fracking) is a booming industry with poorly quantified effects on local air quality.40 Little research to date has explored the contribution of air pollution to CVD in AIs, highlighting another important avenue for future studies.
In AI communities, diet has changed substantially over the past several decades. The malnutrition and undernutrition common among AI youth in the 1960s and 1970s have been replaced by high rates of obesity and CVD.41 Yet despite this growing burden, large gaps persist in nutrition research with AI youth.41 The small number of extant studies suggests that AI youth consume diets high in fats and added sugars and low in fruits and vegetables.41, 42, 43, 44, 45, 46 However, it is challenging to make generalizations from this body of work, because most studies were conducted in single communities43, 45, 46 with small sample sizes,42, 46 using dietary measures of either unknown or poor validity and reliability in youth.43, 44, 45 In addition, collection of dietary data from children and adolescents can face challenges of limited literacy, memory, attention span, and food knowledge.47 For example, some youth might be unable to accurately identify or recall ingredients in multi‐ingredient meals and might therefore under‐ or over‐report intake of particular foods. Among adults, three 24‐hour recalls are considered necessary to obtain a valid measure of “usual diet,” while as many as nine 24‐hour recalls might be needed for adequate assessment in youth.48 Although some studies with youth have tried to compensate for the use of self‐reported dietary measures by recruiting parents or caregivers as proxies, the proxies themselves often do not know what school‐age youth eat throughout the day.47 This limitation is particularly concerning in research with low‐income populations, including many AI communities, where youth may eat 2 or 3 meals per day at school or during after‐school activities.
SHFS participants completed the Block Food Frequency Questionnaire and a supplemental AI foods questionnaire from 2001 to 2003 to ascertain usual dietary intake over the previous year.49 The Block instrument is one of the most widely used food questionnaires, with demonstrated reliability and validity.50, 51, 52, 53 Overall, as shown in Figure, the diet of SHFS participants was poor relative to that of the US general population.54, 55 In National Health and Nutrition Examination Survey data collected from 1988 to 2008, 25% of all‐races respondents who were free of CVD achieved at least 2 of the 5 dietary goals set by the American Heart Association: at least 4.5 cups of fruits and vegetables per day, at least two 3.5‐ounce servings of fish per week, at least 3 servings of whole grains per day, <1500 mg sodium per day, and <36 ounces of sugar‐sweetened beverages per week.56 Only 6% of the SHFS cohort achieved 2 of these 5 goals, while none achieved more than 2 goals, and 65% achieved none at all.54 Results were consistent across both sexes, all ages, and all study sites.
Despite a large body of evidence demonstrating that poor diet elevates CVD risk, programs that provide dietary education and promote better nutrition for AI youth have not led to long‐term improvements in diet or cardiovascular health. Moreover, efforts to increase the availability of healthy foods at grocery stores and schools in underserved communities have had mixed results,42, 57, 58 as access to healthier foods does not always translate into healthier food choices.59 A lack of familiarity with healthy foods, especially their flavors, can deter children from eating them. Children are also particularly vulnerable to marketing campaigns for unhealthy processed foods.60 Such factors must be considered and addressed in developing and implementing interventions to improve diet among AI youth.
Studies of the determinants and correlates of physical activity among AI/AN adults and children, as well as intervention studies to promote physical activity in Native communities, are just as rare as research in the other areas discussed in this review. A number of studies have addressed overweight and obesity in Native people, with physical activity assessed as a covariate.61, 62, 63, 64, 65 Other studies of the correlates of physical activity in AI/ANs have recruited either children or adults, but not both.66, 67, 68, 69, 70 Only 20% to 27% of AI/AN adults are estimated to engage in regular physical activity.71, 72 Data on AI children in Oklahoma show similar trends, with only 30% meeting the 2008 Physical Activity Guidelines for Americans.73
A systematic literature review of physical activity levels among AI/AN and First Nations people found that 26.5% of children met World Health Organization or Canadian physical activity guidelines according to studies using self‐reported data, while 45.7% of children met guidelines according to studies based on accelerometry or pedometry, although self‐reporting in children may be less reliable.74 Other research shows that AI/AN children who do not participate in team sports, rarely engage in physical activity, and perform poorly on fitness tests have a higher risk of obesity.61, 63 In a study of AI children ages 3 to 8 years in Wisconsin, number of hours of outdoor play significantly predicted percentage of body fat after controlling for age, sex, and maternal body mass index (BMI).75 In the Pathways study with AI children, accelerometry data revealed an association between higher levels of physical activity and smaller percentages of body fat in normal‐weight but not overweight second graders.65
One systematic review identified 33 physical activity interventions in US tribal communities published between 1970 and 2015. Twenty were research‐based interventions and 13 were community‐wide initiatives.76 None reported statistically significant improvements in weight‐related outcomes, and only 4 reported improvements in knowledge or attitudes regarding physical activity.76 The community‐wide studies did not share systematic information or analysis on the local environment, policy, or system‐based approaches.
This body of research highlights the importance of objective measures of physical activity in AI/AN children, especially accelerometry.65, 75, 77 It also emphasizes the need for studies, not only to understand the individual determinants of physical activity in AI/AN youth, but also to develop effective, multilevel strategies to promote physical activity in families and communities through interventions in behavior, policy, health systems, and natural as well as built environments.
Obesity rates in AI/AN children are among the highest of all races and ethnicities.78, 79, 80 The Centers for Disease Control and Prevention estimate that AI/ANs ages 2 to 5 years have a higher combined prevalence of overweight and obesity (58.8%) than children of any other race or ethnicity (30%).78, 81 AI/AN children are also the only group of children for which an increase in obesity prevalence has been reported since 2003.82 Once established, obesity persists into later life and greatly increases the risk of chronic disease.83, 84 In AI communities in Wisconsin, 46% of children aged 3 to 8 years were overweight or obese, with obesity defined as sex‐specific BMI for age >95th percentile.57, 85 AI preschoolers had more than double the prevalence of obesity when compared with the general population and are the only group to experience an increase since 2003.57, 86 Furthermore, 38% of obese children in this study had significantly abnormal blood lipids, higher blood glucose, and higher abdominal adiposity relative to normal‐weight children (P<0.01). Unfortunately, AI parents and primary caregivers rarely recognize overweight in their children and rarely realize that excess weight has adverse health consequences later in life.85 Given evidence that disparities in obesity prevalence emerge in early childhood,87 tailored intervention strategies at a range of intensities are needed for young AI/AN children.
Research is also needed to understand the effects of prenatal and early postnatal factors on adiposity, cardiovascular risk markers, and the trajectory of BMI in AI/ANs. In WINGS (Wisconsin Indian Nutrition and Growth Study), 471 AI children aged 5 to 8 years from 3 Wisconsin tribes were assessed with anthropometric measures, body fat measures, and nonfasting lipids and glucose by fingerstick blood samples.88 Tribal records from Women, Infants, and Children programs and clinic charts provided longitudinal data on the trajectory of children's BMI, maternal prenatal factors, and the early postnatal feeding environment. Forty‐seven percent of children were overweight or obese. Analysis of growth trajectories showed that children's BMI category was largely determined within the first year of life. Significant predictors of BMI at age 1 year included macrosomia (odds ratio [OR]: 4.38), excess gestational weight gain (OR: 1.64), and early termination of breastfeeding (OR: 1.66). Children who were overweight or obese at age 1 year had greater odds of being overweight (OR: 3.42) or obese (OR: 3.36) and having unhealthy levels of body fat (OR: 2.95) and low‐density lipoprotein (OR: 1.64) at ages 5 to 8 years.
Despite this evidence for the importance of the first years of life in individual CVD risk, few primary prevention studies have targeted young children.89 Yet early childhood is a critical window for the development of health‐related behaviors, such as food preferences and physical activity, with guidance from parents and caregivers.90 Family‐level interventions are therefore urgently needed to encourage healthy eating and physical activity among AI/AN children.
While there are no systematic data on cardiovascular disease among the AN pediatric population, data in younger adults show prevalent CVD risk factors such as obesity and impaired glucose metabolism that provide background information to use in planning studies and interventions. Older literature generally assumed that ANs were protected against CVD by lifestyle factors, although a comprehensive review in 2003 found little evidence to support this hypothesis. The authors reported that incident CVD has been increasing among ANs at least since the 1980s, and that current incidence rates for hypertension and CVD are similar to those in the all‐races population of Alaska and the United States overall.91 They related the increase in CVD to a high community prevalence of cigarette smoking, obesity, and diabetes mellitus; ongoing lifestyle changes that favored physical inactivity; and consumption of nonmarine fatty acids.91 Further, CVD mortality in ANs appears to be comparable to that of the general population.92
Members of our author group recently published a review of the literature on CVD in ANs that has appeared since 2013.93 We draw on that review in the following paragraphs.
The Alaska Siberia Project, a cross‐sectional survey conducted in 1994, was one of the earliest epidemiologic studies to use thorough and validated methods to assess prevalent CVD and its risk factors in ANs.94 This project was modeled closely on SHS.95 Investigators recruited more than 50% of residents aged 25 years or older of 4 villages in the Bering Straits region of Northwestern Alaska. A subsequent study funded by the National Heart, Lung, and Blood Institute, the GOCADAN (Genetics of Coronary Artery Disease in Alaska Natives), investigated genetic and nongenetic determinants of CVD and its risk factors in a sample of 1214 predominantly Inupiat participants from the Norton Sound region of Northwestern Alaska. Baseline data collection was conducted from 2000 to 2004, with follow‐up in 2006 to 2010.96 The EARTH (Education and Research towards Health) study, funded by the National Cancer Institute, recruited AN participants from several regions, including more than 1000 from the Yukon‐Kuskokwim Delta, for a study of chronic diseases (including cancer) and their risk factors.97 The CANHR (Center for Alaska Native Health Research), funded by the National Center for Research Resources, started in the early 2000s and focused on risk and protective factors for obesity and chronic diseases, including CVD.98, 99 CANHR obtained biological, genetic, nutritional, and psychosocial data from examination of nearly 1000 ANs aged 14 years or older from 10 communities in the Yukon‐Kuskokwim Delta.98, 99 However, they report the cardiometabolic baseline profile of those aged 18 years or older (n=753) and found that CANHR participants with normal glycemia were younger and had a healthier cardiometabolic profile compared with those with impaired fasting glucose or with diabetes mellitus.99
The WATCH study (Western Alaska Tribal Collaborative for Health) is a unique collaboration among investigators with CANHR, GOCADAN, and EARTH. Its purpose is to harmonize data and methods in order to improve surveillance of CVD and diabetes mellitus and to facilitate studies of their risk factors and outcomes in ANs.100 The WATCH cohort comprises 4569 western AN people (2116 men, 2453 women). WATCH investigators collect prospective surveillance data with an average 6.7‐year follow‐up on a subset of the cohort comprising 2754 participants (1218 men, 1536 women) who consented to continued surveillance. They have established that total mortality rates among western AN peoples are higher than those of US whites.101 CVD was the leading cause of death, with mortality rates only slightly higher in men than in women. Although mortality rates attributable to coronary heart disease tended to be lower among WATCH participants than among US whites, rates of fatal stroke appeared to exceed those in US whites. Among WATCH women, mortality rates because of stroke were almost as high as those for coronary heart disease. Among both sexes, the prevalence of hypertension and prehypertension was slightly lower than in the US general population, while the prevalence of diabetes mellitus was lower not only than in the general population but also than in AI/ANs residing in the contiguous 48 states.102
Measures of subclinical disease confirm that the high prevalence of CVD risk factors among AI/ANs leads to significant adverse changes in cardiac structure and function. Combined with other findings in the literature, this conclusion indicates a high probability that the incidence of CVD events in Native adults will continue to rise. Notably, AI/AN youth with multiple risk factors exhibit more extensive subclinical CVD than those without.89, 91 Thus it seems obvious that atherosclerosis starts in childhood and is associated with the same risk factors as CVD in adults. Accordingly, more data on CVD risk factors unique to AI/AN youth are sorely needed. In addition, a glaring omission in the medical literature is an understanding of adverse childhood experiences and other psychosocial factors and their relation to CVD in AI/ANs. There is a fair body of literature associating adverse childhood experiences with CVD in the general population, though our knowledge of adverse childhood experiences in indigenous populations is limited to their prevalence in AI/ANs.103 The prevalence of a significant adverse childhood experience score (3 or more) is more than double that of the general population, which highlights the need to explore the long‐term effects of this disparity.104 Furthermore, we identify exposure to environmental contaminants, diet, and physical activity as areas in particular need of future research.
Few studies have assessed the role of exposure to metals and environmental pollutants in the development of CVD in AI/ANs. Major population disparities are known to exist in exposure to environmental toxicants, and AI communities in particular experience heightened risk. Unfortunately, environmental exposures are involuntary and often beyond the control of exposed individuals. Moreover, they are likely to have adverse health effects not only in childhood but even before birth, potentially before other pathophysiological risk factors for CVD become significant. For these reasons, cohort studies are needed to understand the contribution of environmental exposures to incident CVD.
The challenges of collecting dietary data from youth and the paucity of longitudinal studies among AI/ANs are well known. Nevertheless, it is clear that the diets of Native youth, like those of most Americans, are lacking in fresh fruits, vegetables, meats, and grains, with disproportionate contributions from processed foods and sugar‐sweetened beverages. Early dietary interventions are critical, as life‐shaping habits develop early in life, while the risk of obesity and other cardiometabolic disease increases with age. Although some intervention research has promoted healthy eating among AI/AN youth, most programs are brief, and they have shown little efficacy in achieving long‐term dietary improvement. Negative outcomes might be attributable in part to an overly narrow scope of implementation (eg, school‐based programs with little effect outside the classroom). Thus, multilevel interventions that encompass children, families, and communities are essential, as are sustainable programs that can be implemented in resource‐limited settings with inadequate access to healthy foods.
Our understanding of the role of physical activity in preventing CVD among AI/AN youth is restricted by the paucity of randomized trials designed to increase physical activity in community settings. With regard to obesity, few interventions have prioritized adolescents, and fewer still have targeted prenatal or early‐life influences. With regard to emerging CVD risk factors, such as stress, sleep, nutritional factors, seasonal change, climate change, and changes in the gut microbiome, the state of research is even more rudimentary. Investigators urgently need to understand which improvements in policies, systems, or physical environments can promote and sustain healthy behaviors. Studies of community‐level policies and systems are ill‐suited to randomized trials, given the small size of many AI/AN communities and the expense of community‐based programs, yet such trials are essential.
With regard to ANs, further pediatric research and engagement is needed on novel as well as traditional CVD risk factors, particularly among young children. Continued surveillance is indispensable to elucidate emerging sources of risk, including genetic and environmental factors. Studies must also examine lifestyle factors (eg, traditional patterns of diet and physical activity) that might explain community‐level differences in diabetes mellitus incidence. Also needed are educational interventions that use community resources and highlight resiliency in reducing CVD risk.
Thankfully, our knowledge of CVD risk factors in AI/AN youth is expanding. Noninvasive imaging has established that children and adolescents with multiple CVD risk factors also have subclinical CVD. Thus, given the notable risk of early CVD mortality among AI/ANs, primary prevention and treatment must begin early in life. Interventions that wait until adulthood have clearly failed to stem the tide of CVD in Native communities. Of course, identifying pediatric predictors of incident CVD will require focused longitudinal study, but we believe that cardiovascular screening during pediatric care is one of the most promising avenues for reducing the harsh disparities in CVD mortality experienced by Native people. This strategy might ultimately diminish the burden of CVD in a highly vulnerable US population.
The authors would like to acknowledge the forum planning committee, particularly Mollie A. Minear, PhD and Mona Puggal, MPH as well as Raymond M. Harris, PhD for his thoughtful edits of this article. In addition, we thank the Center for American Indian and Rural Health Equity, Montana State University, NIGMS 5P20GM104417, for funding the article publication costs.
J Am Heart Assoc. 2017;6:e007576 DOI: 10.1161/JAHA.117.007576.