Abstract

Blacks and Latinos are less likely than whites to access health insurance and utilize health care. One way to overcome some of these racial barriers to health equity may be through advances in technology that allow people to access and utilize health care in innovative ways. Yet, little research has focused on whether the racial gap that exists for health care utilization also exists for accessing health information online and through mobile technologies. Using data from the Health Information National Trends Survey (HINTS), we examine racial differences in obtaining health information online via mobile devices. We find that blacks and Latinos are more likely to trust online newspapers to get health information than whites. Minorities who have access to a mobile device are more likely to rely on the Internet for health information in a time of strong need. Federally insured individuals who are connected to mobile devices have the highest probability of reliance on the Internet as a go-to source of health information. We conclude by discussing the importance of mobile technologies for health policy, particularly related to developing health literacy, improving health outcomes, and contributing to reducing health disparities by race and health insurance status.

Blacks and Latinos are more likely to live in poor neighborhoods, have higher rates of mortality, and suffer from chronic diseases (Gilbert et al. 2016; Gilbert, Ray, and Langston 2014; Massey and Denton 1993; Sewell 2010). Wealth and economic segregation is important when thinking about racial differences in health outcomes, considering that the high level of obesity among blacks is explained away by their likelihood of living in poverty (Boardman et al. 2005; Diez-Roux 2001, 2011; Hajat et al. 2010; Kawachi and Berkman 2003). As a result, scholars consider racial residential segregation as a fundamental cause of health disparities (Kwate et al. 2009; Williams and Collins 1995).

In addition to housing segregation, blacks and Latinos face labor market discrimination that restricts their access from jobs that provide quality health insurance (Stainback and Tomaskovic-Devey 2012). Consequently, blacks and Latinos are less likely to utilize health care and face a daunting health insurance gap. As of 2014, nearly 30 percent of adults who are Latino or living in poverty, as well as over 15 percent of adult blacks, compared to around 10 percent of adult whites, did not have health insurance coverage.1 Besides the insurance gap, blacks are less likely to trust physicians (Sewell 2015). Due to a lack of communication between minority patients and health care providers, compared to white patients, and the fact that health care personnel mistakenly perceive blacks as having higher pain thresholds, minorities often receive fewer services and lower quality care (Nelson, Stith, and Smedley 2003). Thus, even when blacks and Latinos have adequate health insurance, they still face barriers to high-quality care in the patient-provider relationship.

One way to overcome some of these racial barriers to health care equity is through advances in technology that allow people to utilize health care in new ways. Yet, little research has focused on whether the racial gap that exists for health care utilization also exists for accessing health information online and through mobile technologies. Recent research suggests that the racial technology gap may be less prominent once mobile technologies are examined (Ray, Gilbert, and Sewell 2016). Missing, however, is whether mobile technologies can help provide access to health information to reduce health disparities.

In this article, we use data from the Health Information National Trends Survey (HINTS) to examine racial differences in obtaining health information online via mobile devices. We first provide an overview of research on the racial gap in health care access and utilization. Second, we discuss why mobile technologies may be useful for bridging the racial gap in health care access. Third, we describe our data and methods. Then, we present our findings and conclude by discussing the policy implications our findings have for federally funded programs such as Medicaid, the Obama Phone Program,2 and the Healthy Food Financing Initiative Bill.

Background

The Racial Digital Divide: Issues of Access

“Segregation concentrates economic disadvantage with racial disadvantage” (Ray, Gilbert, and Sewell 2016). Roughly 8 million Americans live in extreme poverty neighborhoods, with 40 percent of the local residents living below the poverty line. Due to housing segregation and discrimination, blacks and Latinos are more likely to live in segregated, poor neighborhoods (Charles 2003). Poor neighborhoods are often isolated from technological infrastructure that becomes useful for taking advantage of advances in health care. Mobile technologies allow for the use of health devices that require wireless technology (Wiehe et al. 2008). For example, pacemakers have wireless capabilities that alert physicians when a patient's heart goes into atrial fibrillation. People who live in a poor neighborhood without access to wireless Internet in their homes cannot take advantage of this technological advancement to monitor their arrhythmias and have a higher likelihood of stroke and heart disease (Go et al. 2001).

Thus, in many ways, racial segregation operates as a structural barrier to health information. Structural barriers constrain the formation of interpersonal relationships with other residents and local and national organizations that connect local residents with trusted health care providers. Consequently, the diffusion of health information into poor and minority communities that can increase health care utilization by taking advantage of cutting-edge policy prescriptions is inhibited (Szreter and Woolcock 2004). This is important considering research that shows that blacks and Latinos are more likely to seek neighborhood resources than individuals in more affluent neighborhoods (Barnes 2003, 2004). This line of research suggests that access is key rather than individual motivation.

Mobile Technologies to Circumvent Issues of Health Information Access

Mobile technologies are different from e-health technologies and may lend themselves to different groups across race and social class in nuanced ways that need to be assessed. Recent research shows that blacks and Latinos may be less likely to have a landline but more likely to have a mobile phone (Ray, Gilbert, and Sewell 2016). They also may be less likely to have home-based Internet but more likely to use the Internet on their mobile phones than whites. How and/or for what purposes people use mobile phone technologies are important questions that have yet to be addressed by extant research. Although many individuals may use mobile technologies for work and leisure, those who are cut off from broadband services may use these technologies for accessing government services, identifying places to deal with food insecurity, and searching the Internet for health information to circumvent the lack of access to health care services and information from health care professionals.

We propose that mobile technologies can serve as central nodes in a network to connect poor, minority, or rural individuals with information that can improve their health. For example, there is an abundance of research on mobile technology applications used to increase physical activity and reduce obesity (Cavallo et al. 2012; Keller et al. 2014; Tate, Jackvony, and Wing 2003, 2004). We believe giving individuals similar access to health information via mobile technologies is vital considering that over 40 percent of US households do not have landline phones (Blumberg and Luke 2014) and instead only have a mobile phone. However, having a mobile phone does not mean that a person has access to the Internet with that mobile phone. A sizable percentage of the US population, who are also more likely to be poor and minority, do not have access to online resources in the same way as others.3

Bridging the Access and Health Information Divide

So, how do mobile technologies help reduce racial health disparities? First, mobile technologies allow health care providers to maintain regular contact with their patients. Similarly, mobile technologies provide researchers a low-cost opportunity to keep participants informed and engaged about research studies (Ray, Gilbert, and Sewell 2016). Second, mobile technologies allow for more real-time health information to be gathered and used. For example, patients with diabetes can take their blood sugar and enter it into a mobile technology application to receive how much insulin to administer.4 This is important considering that the US Pharmacopeia Medication Errors Reporting Program finds that roughly “50 percent of all medication errors involve insulin.”5 Half of these errors occur among individuals who are 60 years of age or older. This program reports that these errors add nearly $2,000 to a patient's total health care costs. As of 2014, over 29 million Americans had diabetes.6 Without the use of these technologies, individuals on insulin continue to run the risk of giving themselves the wrong amounts of insulin. These errors can have grave implications for the patient by increasing morbidity and mortality, but also for state and federal budgets if the patient is on Medicaid or Medicare. This example illuminates how mobile technologies can increase communication between patients and health care providers, reduce costs, and improve overall health.

Third, mobile technologies may allow for patients with limited mobility and trouble accomplishing activities of daily living to interact with their health care providers without having to leave their homes (Vandelanotte et al. 2007). This type of technology is already being used for new mothers who have C-sections and may be limited in their ability to travel in the first month after birth. These technologies also decrease health care costs by reducing the number of days that patients may stay in the hospital. Furthermore, this type of mobile technology is being used by residents in Hawaii who may live on a different island from their health care provider. These types of technological advancements would be very useful for poor residents who have access issues due to limited public transportation routes or distance from health care providers.

With evidence documenting that blacks and Latinos are more likely to use mobile phone technology, coupled with individuals in economically disadvantaged neighborhoods aiming to access more social capital resources (Barnes 2003), there is reason to believe that blacks and Latinos may also use mobile technologies to access health information. We examine whether the digital divide exists for mobile technologies and whether such technologies lead individuals across racial and class divides to access health information in similar ways. While we do understand that accessing health information online is different from interacting with health care providers via mobile technologies, this examination is still fruitful as it may hold insights into ways to further circumvent racial disparities in health care access.

Methods

Data

We analyzed data from the Health Information National Trends Survey (HINTS) 4, Cycle 3. A cross-sectional, mail-administered survey, HINTS is designed to collect nationally representative data about the US population's health communication practices. Data were collected between September and December of 2013 with a response rate of 35.2 percent. A total of 3,165 respondents completed the portion of the survey that included the measures of interest. The sample was further restricted to include only respondents ages 18 to 59 who provided data on key independent variables (racial group membership, insurance type, mobile device access) and sociodemographic attributes. The age and race/ethnicity restrictions of the data account for the largest reductions in the sample size (roughly 42 percent). Although people over the age of 60 are more likely to face chronic diseases than the age group in our study, we assert that our study has the potential to help policy makers get ahead of ways to increase utilization as people age. Given that the study focused on multiple outcomes, the sample size varied slightly based on the outcome of interest. A total of 1,036 respondents provided valid data on all measures of interest.

Measures

Dependent Variables

The primary dependent variables were ascertained from a series of questions concerning attitudes and behaviors toward the sources of health information. Trust in sources of health information was ascertained from the following question: “In general, how much would you trust information about health or medical topics from [source]?” Twelve sources of health information were evaluated: (1) doctor; (2) family or friends; (3) online newspapers; (4) print newspapers; (5) special health or medical magazines or newsletters; (6) radio; (7) internet; (8) local TV; (9) national or cable TV news programs; (10) government health agencies; (11) charitable organizations; and (12) religious organizations and leaders. The four response options included: not at all; a little; somewhat; very. Higher responses indicate more trust. In some analysis, the values were recoded so that higher values indicated trust in the source of health information.

Among people who reported having “ever looked for information about health or medical topics from any source,” recent source of health information was ascertained from the following question: “The most recent time you looked for information about health or medical topics, where did you go first?” The response options were: (1) doctor or health care provider; (2) internet; (3) other (books; brochures, pamphlets, etc.; cancer organization; family; friend/co-worker; library; magazines; newspapers; telephone information number; complementary, alternative, or unconventional practitioner; other). A fourth response option was allowed for people who had not recently sought health information. A four-category outcome was created, as follows: (1) doctor or health care provider; (2) internet; (3) other (list); (4) did not seek health information recently.

Go-to source of health information was ascertained from the following question: “Imagine that you had a strong need to get information about health or medical topics. Where would you go first?” The response options were: (1) doctor or health care provider; (2) internet; (3) other (books; brochures, pamphlets, etc.; cancer organization; family; friend/co-worker; library; magazines; newspapers; telephone information number; complementary, alternative, or unconventional practitioner; other). A binary indicator of reliance on the Internet was also created by coding internet as 1 and all other non-missing responses as 0.

Table 1 provides a description of the variables used in the primary analysis based on the sample of individuals with non-missing data on all outcomes and covariates of interest. For convenience sake, trusting attitudes were dichotomized such that 1 = very trusting; 0 = not at all, a little, or somewhat trusting. All other measures are reported in accordance with how they are used in the analysis.

Independent Variables

The analysis focused on explaining variation in the outcomes of interest due to three factors: (1) mobile device access; (2) insurance type; and (3) racial group membership. For mobile devices, we created a three-category ordinal measure to proxy mobile device access, where: (1) I do not have any of the above (“no mobile device”); (2) tablet computer only, smartphone only, or cell phone only (“1 mobile device”); and (3) multiple devices selected (“2+ mobile devices”). Mobile devices exclude desktop computers and laptops, but do include: tablets like an iPad, Samsung Galaxy, Motorola Xoom, or Kindle Fire; smartphones such as an iPhone, Android, Blackberry, or Windows Phone; and cell phones. This measure captured the relative extent to which a respondent was connected to the Internet via mobile devices. Higher values indicate more ways in which an individual can access the Internet.

For insurance type, we created a three-category nominal variable, where (1) no health care insurance; (2) Medicare, Medicaid, or any other; or (3) employer-provided or insurance purchased directly from an insurance company. All other responses were coded to missing. Of those that were missing, forty-nine respondents did not provide data on health care coverage, and seventy-six respondents reported some other form of health care insurance coverage than ascertained in this study (e.g., VA). Racial group membership was a self-reported measure and includes those who identify as either white, black or African American, or Hispanic/Latino. Due to small sample sizes for other groups, they were excluded from the analysis.

Control Variables

Control variables included sociodemographic attributes and health. Sociodemographic attributes included gender, age, marital status, household size, household income (mid-repoint recode of nine-category ordinal variable with the following categories: $0 to $9,999; $10,000 to $14,999; $15,000 to $19,999; $20,000 to $34,999; $35,000 to $49,999; $50,000 to $74,999; $75,000 to $99,999; $100,000 to $199,999; $200,000 or more), highest educational level in household, and occupational status. Health indicators included a five-category ordinal measure of feeling sad in the most recent month, lifetime cancer diagnosis, and self-rated health.

Analytic Strategy

We noted missing data for both the dependent and independent variables for the 18–59 population. Percentages with missing data for the dependent variables were low, ranging from 0.7 percent (trust in doctors) to 7.4 percent (recent source of health information). Percentages of respondents having missing data on independent variables ranged from 0.0 percent (geographical region, marital status) to 3.5 percent (occupational status). We assumed the data were missing at random, and deleted all cases from the list with missing data on independent variables allowing for variation in sample sizes based on the dependent variable of choice.

All analyses were adjusted for the complex survey design of HINTS. Survey estimation (svy:) commands in Stata 14 were used to include probability weights (person_finwt0), clustering by Designated Market Areas (dma), and three different sampling strata (stratum). Trust in sources of health information outcomes were measured using ordered logistic regression with an allowance for five response categories. For behaviors and attitudes toward the Internet as a source of health information, multinomial logistic regression was used with health care providers as the reference category. Wald tests for combining alternatives (Long and Freese 2014) were conducted to determine if there were differences in behaviors and attitudes toward sources of health information among the different types. These tests suggested that the Internet response option could not be combined with other nonmedical response options or with the Never category (for recent source of health information). The tests also revealed that health care providers could not be combined with other outcome responses. So, we analyzed three categories: Internet, health care providers, and other nonmedical sources.

In the analysis that follows, we provide four types of estimates. First, we include estimates of racial disparities in trusting twelve different sources of health information, holding constant sociodemographic attributes and health-related indicators. Second, we provide estimates of racial disparities in utilization behaviors and willingness attitudes toward the specified sources of health information, holding constant sociodemographic attributes and health-related indicators. Third, we provide estimates of racial disparities in attitudes toward the Internet as a go-to source of health information, while adjusting for the main and moderating effect of insurance type and for the main and moderating effect of mobile device access. Fourth, we provide estimates of racial disparities in attitudes toward the Internet as a go-to source of health information, while adjusting for the ways that insurance type moderates the main effect of mobile device access. For each type of evaluation with second-order interactions, we produced adjusted models that control for sociodemographic attributes, health-related indicators, and trust in the Internet as a source of health information (see figs. 1, 2, and 3).

Limitations

The data in our study obviously have limitations that impact our analysis. First, the survey had a low response rate. Though this response rate is similar to—and, in some regards, better than—other studies that sample using traditional phone methods, we do not have information for the potential respondents who opted out of the study either due to non-response or refusal. Therefore, there may be response bias present in the sample, though research is mixed on whether non-respondents differ from respondents in nationally representative and/or random samples (Abraham, Helms, and Presser 2009; Tourangeau, Groves, and Redline 2010). Second, we were unable to ascertain the specific types of websites and news sources that respondents obtained from the Internet. Third, it is possible that people who use their mobile devices to do Internet searches for health information may have unmeasured characteristics that also are associated with getting mobile devices in the first place, leading to endogeneity issues (e.g., doing Internet searches may make them care more about accessing health information than people who do not do those things). Fourth, although we purport that accessing health information online may translate to telehealth opportunities, assessing this potential payoff is beyond the scope of this study. We also do not want to imply that telehealth can be used as a substitute for direct interaction with health care providers. Rather, we view telehealth as another tool for health care providers to use, and the potential payoff from these opportunities may benefit individuals living in impoverished communities. Future research should aim to explore these possibilities given policy changes that aim to enhance wireless capabilities.

Results

Racial Disparities in Trusting Sources of Health Information

Table 2 shows racial disparities in trusting twelve different sources of health information. The estimates adjust for racial differences in sociodemographic attributes that may also be correlates of trust. The analysis reveals heterogeneity in the nature and degree of black-white and Latino-white disparities in trust in sources of health information.

Blacks are more likely to trust health information from online news sources, radio, local TV, national TV, and religious organizations than whites, while Latinos are more likely to trust online news, local TV, national TV, and the government than whites. Latinos are less likely than whites to trust health information from family members; yet, no black-white differences exist in trusting this source of health information. In addition to there being no racial differences in trusting doctors for health information, there are also no racial differences in trusting print news, health news, the Internet, and charities. There are also no statistically significant black-white differences in trusting government, and no Latino-white differences in trusting religious organizations.

Racial Disparities in Use of and Reliance on the Internet for Health Information

Table 3 shows variation in recent use of the Internet and reliance on the Internet in times of strong need by racial group membership, insurance type, and mobile device access status—the key covariates of interest in this study. The results reveal that distrust in the Internet is associated with lack of use and reliance on the Internet as a source of health information. The estimates indicate that people who think the Internet is not a trustworthy source of health information are less likely to actually have used the Internet in the most recent search for health information compared to using a doctor. They are also less likely to use the Internet as a go-to source for health information in times of strong need for information about health.

Holding constant distrust in the Internet, there are no racial differences in recent use of the Internet, but blacks are less likely than their white counterparts to rely on the Internet and other nonmedical sources. That is, blacks are more likely to report preferring to go to a doctor or health care provider during times of strong need. There are no Latino-white differences in either use of or reliance on the Internet.

There are, however, gaps in Internet use and reliance based on insurance type. Individuals who are insured by Medicare or Medicaid (i.e., federally insured) are less likely than the uninsured to use the Internet when looking for health information recently. In fact, insured persons (both federally and privately insured) are less likely than the uninsured to indicate the Internet as a go-to source of health information in times of strong need. There are no main associations with mobile device status, holding constant these key covariates of attitudes toward the Internet.

Differential Reliance on the Internet by Insurance Type across Racial Groups

Uninsured individuals of all races are more likely to rely on the Internet as their go-to source for health information. However, the gap between the uninsured and the privately insured is smaller for blacks and Latinos than among whites. Compared to whites, the uninsured/privately insured gap is 5.92 times smaller among blacks (z = 2.74) and 3.65 times smaller among Latinos (z = 2.54) (table 4, column 1). These race differentials in the relative association of insurance type with reliance on the Internet hold when adjusting for sociodemographic attributes (fig. 1; table 4, column 2).

Differential Reliance on the Internet by Mobile Device Access across Racial Groups

Having a mobile device is associated with reliance on the Internet as a go-to source of health information for all racial groups. However, the relative extent to which the mobile device divide matters for Internet health information reliance varies by racial group. For instance, there is a smaller gap between people who have two or more mobile devices and people who have no mobile device among whites than among either blacks or Latinos (table 4, column 3). Blacks who do not have a mobile device face a serious dearth of reliance on the Internet. Virtually no blacks in these data who did not have a mobile device reported that they would turn to the Internet for health information in a time of need (fig. 2; table 4, column 4). The probability of Internet reliance of Latinos who are not connected is only slightly higher than the comparable probability for blacks. Even though connected blacks are less likely to rely on the Internet than connected whites, the large magnitude of the mobile device divide between blacks with no devices, compared to one or two devices, indicates that being connected provides blacks with a vastly different orientation toward the Internet and its relationship with health information than is the case among whites.

Differential Reliance on the Internet by Mobile Device across Insurance Types

Having a mobile device is associated with reliance on the Internet as a go-to source of health information for all insurance types. However, the relative extent to which the mobile device divide matters for Internet health information reliance varies by insurance type. For instance, there is a larger gap between people who have two or more devices and people who have no mobile devices among the uninsured and the privately insured than among the federally insured (table 4, column 5). Holding constant sociodemographic attributes, federally insured persons who do not have a mobile device are more likely to rely on the Internet as a go-to source for health information than uninsured and privately insured persons who have one mobile device (fig. 3; table 4, column 6). These findings indicate that the mobile device divide is nearly nonexistent among federally insured persons, but is prominent among people who lack insurance and people who have private insurance. Mobile devices may serve similar functions for the uninsured and the privately insured.7

Health Policy Implications

Establishing health equity for all Americans regardless of race/ethnicity or social class background is not solely about economic resources. Rather, health equity is also about social resources and providing individuals with the ability to exert agency to maintain and improve their health outcomes. Providing access to health information allows people to obtain knowledge important to their well-being more expeditiously. Health information dissemination improves health literacy and may enhance compliance and behavioral change. It also allows people to become more prepared for interactions with health care providers and feel greater autonomy over their lives. For people with chronic health conditions such as diabetes, heart failure, and kidney failure, mobile and wireless technologies allow them to maintain care in outpatient settings rather than in hospital settings. This change in setting helps to reduce health care costs for insurance companies and government-sponsored insurance programs.

Despite these benefits, poor and minority community members are less likely to have access to mobile and wireless technology, yet more likely to have chronic health conditions. Providing wireless infrastructures to improve telehealth capabilities in poor and minority communities could help to overcome the lack of health care providers and inadequate resources in such neighborhoods. With these infrastructures, mobile technologies may be used to document how travel patterns and neighborhood locations are associated with health-risk behaviors across the life course of youth and aging adults. Nonetheless, providing access to mobile and wireless technologies is not a substitute for dealing with the real challenges of building an infrastructure of health that provides equitable opportunities for health care. Moreover, access to health information via mobile technology is simply a complement to primary care services for the underserved. Strategies that provide ways to enhance patient agency, while also aiming to overhaul an inequitable health care system, should be pursued simultaneously.

One health policy program, in particular, may lend itself to leveraging mobile technology for poor Americans. The Obama Phone Program allows residents who meet a certain financial threshold to obtain a mobile phone to communicate with employers, job prospects, and family members and friends. These phones, however, only give a person the ability to call and text. What is needed is the ability to gain access to the Internet. As we show here, blacks and Latinos will definitely heed these opportunities to access the Internet in order to obtain vital health information. Incentives and subsidies can be offered to corporations that provide wireless infrastructure in poor neighborhoods. This is in line with Healthy People 2010 objectives that aimed to increase the number of people with home-based Internet.

Altogether, mobile technologies are vital for health care as we progress into the twenty-first century. Finding ways to enhance existing programs in order to provide access to mobile technologies will correspondingly improve health literacy and outcomes. Mobile technologies may also contribute to reducing racial health disparities among whites, Latinos, and blacks, as well as those individuals on Medicare and Medicaid.

Conclusion

This article examined racial differences in the propensity to access health information online. Blacks and Latinos, compared to whites, were more likely to trust online newspapers to get health information. Blacks also were more likely than whites to use the Internet to access health information when in the midst of a strong need event. However, minorities who are privately insured were more likely than their uninsured counterparts to rely on the Internet. These findings are important considering that federally insured persons who are connected to mobile devices had the highest probability of reliance on the Internet as a go-to source of health information. In sum, these findings lend credence that mobile technologies are important for achieving greater racial equity in health behavior and health outcomes.

Acknowledgments

The authors acknowledge the JHPPL special issue editors and reviewers as well as participants of the 2016 Robert Wood Johnson Foundation Investigator Awards conference for helpful comments on this manuscript.

Funding for this project was provided by the Robert Wood Johnson Foundation Health Policy Research Program.

Notes

4. This information was obtained from Vanderbilt University radiologist Dr. Laveil Allen.

7. Supplemental analysis (available upon request) examined the probability of a three-way interaction between racial group membership, mobile device access, and insurance type; however, sample sizes for minorities were too small to conduct a reliable test of third-order interactions.

References

References
Abraham
Katherine G.
,
Helms
Sara
, and
Presser
Stanley
.
2009
. “
How Social Processes Distort Measurement: The Impact of Survey Nonresponse on Estimates of Volunteer Work in the United States
.”
American Journal of Sociology
114
:
1129
65
.
Barnes
Sandra L.
2003
. “
Determinants of Individual Neighborhood Ties and Social Resources in Poor Urban Neighborhoods
.”
Sociological Spectrum
23
, no.
4
:
463
97
.
Barnes
Sandra L.
2004
. “
Too Poor to Get Sick? The Implications of Place, Race, and Costs on the Healthcare Experiences of Residents in Poor Urban Neighborhoods
.”
Chronic Care, Healthcare Systems and Services Integration: Research in the Sociology of Healthcare
22
:
47
64
.
Blumberg
Stephen J.
, and
Luke
Julian V.
2014
. “
Wireless Substitution: Early Release of Estimates from the National Health Interview Survey, January–June 2014
.”
US Department of Health and Human Services. Centers for Disease Control and Prevention. National Center for Health Statistics
.
1
14
.
Boardman
Jason D.
,
Saint Onge
Jarron M.
,
Rogers
Richard G.
, and
Denney
Justin T.
2005
. “
Race Differentials in Obesity: The Impact of Place
.”
Journal of Health and Social Behavior
46
:
229
43
.
Cavallo
David N.
,
Tate
Deborah F.
,
Ries
Amy V.
,
Brown
Jane D.
,
Devellis
Robert F.
, and
Ammerman
Alice S.
2012
. “
A Social Media-Based Physical Activity Intervention: A Randomized Controlled Trial
.”
American Journal of Preventive Medicine
43
, no
5
:
527
32
.
Charles
Camille Z.
2003
. “
The Dynamics of Racial Residential Segregation
.”
Annual Review of Sociology
29
:
167
207
.
Diez-Roux
A. V.
2001
. “
Investigating Neighborhood and Area Effects on Health
.”
American Journal of Public Health
91
:
1783
89
.
Diez-Roux
Ana V.
2011
. “
Complex Systems Thinking and Current Impasses in Health Disparities Research
.”
American Journal of Public Health
101
:
1627
34
.
Gilbert
Keon L.
,
Ray
Rashawn
, and
Langston
Marvin
.
2014
. “
Social Dis(ease) of African American Males and Health
.” In
Urban Ills: Twenty-First-Century Complexities of Urban Living in Global Contexts
, vol.
2
., edited by
Yeakey
Carol Camp
,
Thompson
Vetta Sanders
, and
Wells
Anjanette
,
23
-
36
.
Lanham, MD
:
Lexington Books
.
Gilbert
Keon
,
Ray
Rashawn
,
Siddiqi
Arjumand
,
Griffith
Derek
,
Baker
Elizabeth
,
Shetty
Shivan
, and
Elder
Keith
.
2016
. “
Visible and Invisible Trends in African American Men's Health: Pitfalls and Promises
.”
Annual Review of Public Health
37
:
295
311
.
Go
Alan S.
,
Hylek
Elaine M.
,
Phillips
Kathleen A.
,
Chang
YuChiao
,
Henault
Lori E.
,
Selby
Joe V.
, and
Singer
Daniel E.
2001
. “
Prevalence of Diagnosed Atrial Fibrillation in Adults—National Implications for Rhythm Management and Stroke Prevention: The AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study
.”
Journal of the American Medical Association
285
, no.
18
:
2370
75
.
Hajat
A.
,
Kaufman
J. S.
,
Rose
K. M.
,
Siddiqi
A.
, and
Thomas
J. C.
2010
. “
Do the Wealthy Have a Health Advantage? Cardiovascular Disease Risk Factors and Wealth
.”
Social Science and Medicine
71
:
1935
42
.
Kawachi
Ichiro
, and
Berkman
Lisa F.
2003
.
Neighborhoods and Health
.
New York
:
Oxford University Press
.
Keller
Brett
,
Labrique
Alain
,
Jain
Kriti M.
,
Pekosz
Andrew
, and
Levine
Orin
.
2014
. “
Mind the Gap: Social Media Engagement by Public Health Researchers
.”
Journal of Medical Internet Research
16
, no.
1
:
e8
.
Kwate
Naa Oyo A.
,
Yau
Chun-Yip
,
Loh
Ji-Mengh
, and
Williams
Donya
.
2009
. “
Inequality in Obesigenic Environments: Fast Food Density in New York City
.”
Health and Place
15
:
364
73
.
Long
J. Scott
, and
Freese
Jeremy
.
2014
.
Regression Models for Categorical Dependent Variables Using Stata
.
College Station, TX
:
Stata
.
Massey
Douglas S.
, and
Denton
Nancy
.
1993
.
American Apartheid: Segregation and the Making of the Underclass
.
Cambridge, MA
:
Harvard University Press
.
Nelson
Alan R.
,
Stith
Adrienne Y.
, and
Smedley
Brian D.
, eds.
2002
.
Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care
.
Washington, DC
:
National Academies
.
Ray
Rashawn
,
Gilbert
Keon L.
, and
Sewell
Abigail A.
2016
. “
Mobile Technology as a Conduit for Reducing Obesity-Related Health Disparities
.”
Issues in Race and Society
4
, no.
1
:
98
119
.
Sewell
Abigail A.
2010
. “
A Different Menu: Racial Residential Segregation and the Persistence of Racial Inequality
.” In
Race and Ethnic Relations in the 21st Century: History, Theory, Institutions, and Policy
, edited by
Ray
Rashawn
,
279
90
.
San Diego
:
Cognella Publishing
.
Sewell
Abigail A.
2015
. “
Disaggregating Ethnoracial Disparities in Physician Trust
.”
Social Science Research
54
:
1
20
.
Stainback
Kevin
, and
Tomaskovic-Devey
Donald
.
2012
.
Documenting Desegregation: Racial and Gender Segregation in Private-Sector Employment Since the Civil Rights Act
.
New York
:
Russell Sage Foundation
.
Szreter
Simon
, and
Woolcock
Michael
.
2004
. “
Health by Association? Social Capital, Social Theory, and the Political Economy of Public Health
.”
International Journal of Epidemiology
33
, no.
4
:
650
67
.
Tate
Deborah F.
,
Jackvony
Elizabeth H.
, and
Wing
Rena R.
2004
. “
Effects of Internet Behavioral Counseling on Weight Loss in Adults at Risk for Type 2 Diabetes: A Randomized Trial
.” In
Year Book of Psychiatry and Applied Mental Health
,
96
97
.
Tate
Deborah F.
,
Jackvony
Elizabeth H.
, and
Wing
Rena R.
2003
. “
Effects of Internet Behavioral Counseling on Weight Loss in Adults at Risk for Type 2 Diabetes
.”
Journal of the American Medical Association
289
, no.
14
:
1833
36
.
Tourangeau
Roger
,
Groves
Robert M.
, and
Redline
Cleo D.
2010
. “
Sensitive Topics and Reluctant Respondents: Demonstrating a Link between Nonresponse Bias and Measurement Error
.”
Public Opinion Quarterly
74
:
413
32
.
Vandelanotte
C.
,
Spathonis
K. M.
,
Eakin
E. G.
, and
Owen
N.
2007
. “
Website-Delivered Physical Activity Interventions: A Review of the Literature
.”
American Journal of Preventive Medicine
33
, no.
1
:
54
64
.
Wiehe
Sarah E.
,
Hoch
Shawn C.
,
Liu
Gilbert C.
,
Carroll
Aaron E.
,
Wilson
Jefery S.
, and
Fortenberry
J. Dennis
.
2008
. “
Adolescent Travel Patterns: Pilot Data Indicating Distance from Home Varies by Time of Day and Day of Week
.”
Journal of Adolescent Health
42
:
418
20
.
Williams
David R.
, and
Collins
Chiquita
.
1995
. “
U.S. Socioeconomic and Racial Differences in Health
.”
Annual Review of Sociology
21
:
349
86
.