Skip to main content
NCBI home page
American Journal of Public Health logoLink to American Journal of Public Health
. 2015 Nov;105(11):2359–2365. doi: 10.2105/AJPH.2015.302781

Differences by Sexual Orientation in Expectations About Future Long-Term Care Needs Among Adults 40 to 65 Years Old

Carrie Henning-Smith 1,✉, Gilbert Gonzales 1, Tetyana P Shippee 1
PMCID: PMC4605153  PMID: 26378822

Abstract

Objectives. We examined whether and how lesbian, gay, and bisexual (LGB) adults between 40 and 65 years of age differ from heterosexual adults in long-term care (LTC) expectations.

Methods. Our data were derived from the 2013 National Health Interview Survey. We used ordered logistic regression to compare the odds of expected future use of LTC among LGB (n = 297) and heterosexual (n = 13 120) adults. We also used logistic regression models to assess the odds of expecting to use specific sources of care. All models controlled for key socioeconomic characteristics.

Results. Although LGB adults had greater expectations of needing LTC in the future than their heterosexual counterparts, that association was largely explained by sociodemographic and health differences. After control for these differentials, LGB adults were less likely to expect care from family and more likely to expect to use institutional care in old age.

Conclusions. LGB adults may rely more heavily than heterosexual adults on formal systems of care. As the older population continues to diversify, nursing homes and assisted living facilities should work to ensure safety and culturally sensitive best practices for older LGB groups.

According to some estimates, approximately 4% of American adults aged 18 years or older self-identify as lesbian, gay, or bisexual (LGB),1 including more than 1.5 million LGB adults aged 65 years or older.2 This number is projected to grow to nearly 3 million by 2030.2 Research has shown that LGB individuals report worse physical and mental health outcomes and are more likely to engage in harmful health behaviors than their heterosexual peers.3,4 Public health studies often attribute LGB health disparities to minority stress, or the chronic stress associated with being a member of a marginalized minority group.5–9

Elderly LGB individuals are particularly vulnerable to poor health outcomes owing to minority stress because of their experiences of stigma, discrimination, and violence.10–12 Recent studies involving data from California and Washington State indicate that LGB adults aged 50 years or older are more likely than their heterosexual counterparts to report symptoms of psychological distress, physical disability, and chronic disease; they are also more likely to report that their overall health status is poor.13–15

Older LGB adults in same-gender relationships, especially women, are more likely than older heterosexual adults to need assistance with activities of daily living (ADLs) such as dressing, bathing, and doing errands alone,16 signaling a greater need for long-term care (LTC) services and support in later life.17 Such services and support might include assistance with ADLs (e.g., dressing, bathing, eating), instrumental ADLs (e.g., running errands, managing medications, preparing meals), and clinical or nursing tasks (e.g., pain management, physical or occupational therapy, management of incontinence). LTC may be provided in the home or community setting, by paid or unpaid caregivers, or in institutional settings. Although LGB individuals may have an elevated need for LTC services in later life, we are not aware of any studies examining differences in LTC expectations according to sexual orientation.

Most Americans will need LTC at some point in their lives.18 According to current projections, nearly three quarters of all Americans will use LTC during their lives, and nearly half of Americans aged 65 years or older will spend time in a nursing home.18,19 However, middle-aged Americans have unrealistically low expectations of needing LTC,20 and very few people plan for it by either purchasing LTC insurance or making other advanced arrangements.21 This situation may be partly attributable to the high cost of LTC insurance premiums.22 Another major reason why people may not plan ahead for LTC is that they expect family members or close relatives to step in and provide care should they need it.20 This aligns with current use of unpaid caregivers, with approximately 80% of older adults reporting that family members provide the majority of their noninstitutional care.23

However, older LGB adults may have different family structures than older heterosexual adults1; for example, they are less likely to be married, less likely to have children, and more likely to experience conflict with their family of origin.2,24,25 Older LGB adults, especially men, are more likely than their heterosexual counterparts to live alone,13,15,26,27 which is a major risk factor for both needing LTC and having unmet care needs.28–30

Furthermore, although LGB individuals often exhibit distrust in formal LTC systems,31,32 one study showed that older LGB adults are only half as likely as older heterosexual adults to depend on close relatives for help.2 More than a quarter of LGB older adults report apprehension about discrimination as they age and how it may be manifested in institutional discrimination on the part of health care providers, including LTC providers.33 This should concern practitioners and policymakers given that the older LGB population is growing and our current system of LTC may be inappropriate to meet the needs of this group.

Unfortunately, very little is known about LTC expectations among LGB populations. Research on the general population has shown that LTC expectations and planning behaviors are patterned according to demographic characteristics. Characteristics that may promote LTC planning include older age, female gender, being married, being White, having a college education or above, and having previous experience with LTC.21 Research suggests that LGB older adults are poorer and less financially secure than heterosexual older adults, in part because of limited employment opportunities resulting from institutional and personal discrimination34 and limited (or lack of) ability to receive partner benefits or property inheritances.35 Yet, as noted, there is a dearth of literature examining differences in LTC expectations by sexual orientation.3

The few studies that have examined LTC expectations in older LGB populations indicate that these individuals may use nursing homes sooner than the general population owing to a lack of caregivers at home.2,33 LGB older adults in institutional facilities may be at heightened risk of neglect and abuse as a result of limited knowledge and training among providers and, sometimes, blatant discrimination on the part of staff and fellow residents.36

One survey of LGB older adults and their families indicated that LGB adults were more likely than heterosexual adults to be harassed or mistreated in LTC facilities.37 Altogether, 328 respondents reported 853 instances of mistreatment among LGB older adults in LTC settings.37 Thus, some LGB older adults may avoid nursing homes because of barriers such as fear of discrimination and abuse, concerns about going “back into the closet,” and a reluctance to be separated from their partners.2,32 Instead of using formal LTC settings, some older LGB adults rely on friends and LGB-specific community organizations for assistance in later life.33 Older LGB adults who do reside in institutional settings may keep their sexual orientation a secret or seek out welcoming service providers.

In this study, we sought to fill gaps in existing research by comparing LTC expectations among LGB and heterosexual adults aged 40 to 65 years. Whereas older heterosexual adults may expect to rely on their children and spouses for support in later life, the same may not be true for LGB adults. The findings of this study will be especially important for public health practitioners and policymakers planning for future LTC needs in aging populations.

METHODS

We derived the data for this study from the Integrated Health Interview Series, a harmonized version of the 2013 National Health Interview Survey (NHIS).38 The NHIS, a nationally representative survey of the civilian, noninstitutionalized population of the United States, serves as one of the most comprehensive resources on the health of US residents. The family core questionnaire records basic health and disability information for each household member, and a single random adult in each household is selected for a detailed interview component (the sample adult component) that collects information on health conditions, health behaviors, and access to health care. Our study sample was drawn from the sample adult component in 2013. More than 75% of the selected households completed the survey, and 81% of selected adults completed the sample adult component.39

In 2011, a series of questions on LTC expectations were introduced. Adults aged 40 to 65 years were asked about their perceived likelihood (on a 4-point scale ranging from very unlikely to very likely) of needing help in the future with ADLs such as bathing, dressing, eating, or using the toilet because of a long-term condition. Regardless of how likely respondents believed it was that they would need care, they were also asked about the sources they would rely on to provide help should a need arise. Response options included “my family,” “someone I hire,” “home health care organization,” “nursing home/assisted living,” and “other.” Respondents could select as many options as needed.

Beginning in 2013, a question regarding sexual orientation was added to the sample adult component of the NHIS.40,41 Respondents aged 18 years or older were asked which of the following categories best represented how they thought of themselves: “lesbian or gay”; “straight, that is, not gay”; “bisexual”; “something else”; or “I don’t know the answer.” They also had the option of refusing to answer the question. Although transgender identity was included as a category under “something else,” data on this identity were not released in the 2013 NHIS public use files. We classified respondents as LGB (n = 343) or heterosexual (n = 14 107). We excluded respondents who reported their sexual orientation as something else (n = 26), did not know the answer (n = 71), or refused to answer (n = 111). In addition, to reduce potential “false-positive” reports of sexual orientation, we excluded respondents whose reported sexual orientation did not correspond to the gender of their spouse or partner (n = 28). In doing so, we attempted to minimize errors resulting from misclassification, which can be problematic in small populations.

We used descriptive statistics to characterize the study sample and the Pearson χ2 test to compare characteristics between LGB and heterosexual respondents. To analyze expected need for LTC, we used a series of nested ordered logistic regression models to estimate odds ratios (ORs) for responses in 4 LTC need categories (ranging from very unlikely to very likely) among LGB relative to heterosexual respondents. We subsequently adjusted for relevant sociodemographic characteristics including gender, age, relationship status, race/ethnicity, ratio of family income to the federal poverty threshold (< 100%, 100%–199%, 200%–399%, ≥ 400%), educational attainment (< high school, high school, some college, ≥ college), employment status (employed full or part time vs not employed), region of residence (Northeast, Midwest, South, West), and presence of a minor child in the household.

To determine whether health influenced the relationship between sexual orientation and future likelihood of needing LTC, we conducted nested analyses in which we added health characteristics such as self-reported health status (fair or poor health vs good, very good, or excellent health) and serious psychological distress. Distress was measured with the Kessler-6 scale, which includes 6 questions asking respondents how often in the preceding 4 weeks they have felt hopeless, nervous, restless, sad, worthless, or that everything was an effort. Responses are made on a scale ranging from 0 (none of the time) to 4 (all of the time); scores of 13 or higher indicate serious psychological distress.42

In addition, we assessed whether respondents had any current ADL limitations and whether they had a close family member who, as a result of a chronic illness or disability, had needed help with ADLs for a year or more. Finally, we created logistic regression models to assess the odds of respondents expecting to rely on each source of care after adjustment for sociodemographic and health characteristics.

Stata version 13 (StataCorp LP, College Station, TX) was used in conducting all of the analyses, and survey weights were used to account for the complex sampling design and to generate nationally representative estimates. We report adjusted ORs and 95% confidence intervals (CIs) for each model indicating the association between sexual orientation and LTC expectations. Our final analytic sample included adults aged 40 to 65 years who were not missing information on sexual orientation or expected need for LTC (n = 13 417). Because the amount of missing data for each covariate other than poverty status was less than 5%, we used listwise deletion to account for missing data in our models; in the case of poverty status, NHIS-provided imputation files were used to account for the 8% of missing data. As recommended by the National Center for Health Statistics, we used multiple imputations in Stata (via the “mi” family of commands) to estimate models with imputed data for income.43

RESULTS

Table 1 presents data on the characteristics of the sample. Approximately 2% of the respondents self-identified as LGB. LGB individuals were more likely than their heterosexual counterparts to be male, to be living with a partner, and to never have been married. They were also more likely to be White and well educated and to come from families with higher incomes. In addition, LGB adults were more likely to have a current activity limitation and to have a close family member who had needed LTC in the past. Consistent with previous research, LGB adults were less likely to have children aged 18 years or younger living in the household. There were no significant differences between LGB and heterosexual adults with respect to age, employment status, health status, or psychological distress.

TABLE 1—

Sample Characteristics, by Sexual Orientation: National Health Interview Survey Sample Adult Component, United States, 2013

Sexual Orientation
Characteristic Full Sample (n = 13 417) Heterosexual (n = 13 120) LGB (n = 297) Pa
Participants reporting orientation, % . . . 98 2
Female, % 52 53 44 < .01
Age, y, mean 52.58 52.60 51.72 .05
Marital status, % < .001
 Married 52 53 7
 Living with partner 4 4 21
 Separated, divorced, or widowed 30 30 21
 Never married 14 13 51
Race/ethnicity, % < .01
 White 71 70 79
 Black/African American 13 13 10
 Hispanic 11 12 9
 Asian/all other 5 5 2
Ratio to federal poverty threshold, % < .05
 < 100% 13 13 10
 100%–199% 14 14 13
 200%–399% 34 34 28
 ≥ 400% 39 39 49
Educational attainment, % < .001
 < high school 27 28 15
 High school 9 9 4
 Some college 31 31 34
 ≥ college 33 32 47
Employed, % 66 65 69 .22
Region of residence <.01
 Northeast 17 17 19
 Midwest 23 23 15
 South 38 38 35
 West 22 22 31
Any minor children in household, % 30 30 11 < .001
Fair or poor health, % 17 17 17 .95
Serious psychological distress, % 5 5 7 .1
Activity limitation, % 20 19 26 < .01
Close relative needs long-term care, % 12 12 17 <.05
Open in a new tab

Note. LGB = lesbian, gay, or bisexual.

a

For difference between LGB or bisexual respondents.

As can be seen in Table 2, LGB individuals were more likely than heterosexual individuals to expect to need care in the future. Fifty percent of all LGB respondents stated that they believed it was either likely or very likely that they would need LTC in the future, compared with 39% of heterosexual respondents. There were also significant differences in the specific sources of care individuals would rely on should they need care. In both groups, the most common response was family members; however, whereas three quarters of heterosexual respondents stated that they would turn to family, slightly more than half of LGB respondents reported that they would do so. LGB individuals were more likely than their heterosexual counterparts to say that they would hire someone or use a nursing home or assisted living facility. There were no significant differences by sexual orientation in expectations for relying on multiple sources of care (7% of LGB individuals and 8% of heterosexual individuals expected to use more than 1 source of care).

TABLE 2—

Differences in Long-Term Care Expectations, by Sexual Orientation: National Health Interview Survey Sample Adult Component, United States, 2013

Sexual Orientation
Variable Heterosexual (n = 13 120), % LGB (n = 297), % Pa
Likelihood of needing care someday < .01
 Very unlikely 35 31
 Unlikely 26 19
 Likely 26 28
 Very likely 13 22
Source of careb
 Family 75 53 < .001
 Someone hired 11 19 < .01
 Home health care organization 10 12 .44
 Nursing home/assisted living 10 19 < .001
 Other 5 8 .06
Open in a new tab

Note. LGB = lesbian, gay, or bisexual.

a

For difference between heterosexual and LGB respondents.

b

Respondents could select more than 1 option.

Results from the nested ordered logistic regression models are presented in Table 3. In model 1, self-identification as LGB was associated with higher odds of expected need for LTC in the future (OR = 1.49; 95% CI = 1.07, 2.06; P < .05). After control for demographic characteristics in model 2, the association between LGB status and LTC expectations remained consistent, whereas being employed, living in the Midwest, and having a minor child in the household were all associated with lower odds of expecting to need LTC. Meanwhile, having less than a high school diploma, having an income below 200% of the federal poverty threshold, and being separated, divorced, widowed, or never married were associated with higher odds of expected need for LTC.

TABLE 3—

Odds of Higher Expectations of Needing Long-Term Care: National Health Interview Survey Sample Adult Component, United States, 2013

Characteristic Model 1 (n = 13 417), OR (95% CI) Model 2 (n = 13 280), OR (95% CI) Model 3 (n = 13 112), OR (95% CI)
Lesbian, gay, or bisexual 1.49* (1.07, 2.06) 1.42* (1.01, 1.99) 1.30 (0.92, 1.84)
Female 1.04 (0.97, 1.23) 1.05 (0.97, 1.14)
Age 1.01*** (1.01, 1.02) 1.01*** (1.01, 1.02)
Marital status
 Married (Ref) 1.00 1.00
 Living with partner 1.08 (0.89, 1.30) 0.98 (0.81, 1.18)
 Separated, divorced, or widowed 1.11* (1.02, 1.21) 1.00 (0.91, 1.09)
 Never married 1.14* (1.01, 1.29) 1.10 (0.97, 1.24)
Race/ethnicity
 White (Ref) 1.00 1.00
 Black/African American 0.82*** (0.73, 0.91) 0.84** (0.75, 0.94)
 Hispanic 0.88* (0.77, 1.00) 0.97 (0.86, 1.10)
 Asian/all other 0.95 (0.80, 1.12) 1.00 (0.85, 1.18)
Ratio to federal poverty threshold
 ≥ 400% (Ref) 1.00 1.00
 < 100% 1.41*** (1.23, 1.62) 0.99 (0.86, 1.14)
 100%–199% 1.37*** (1.22, 1.53) 1.09 (0.97, 1.22)
 200%–399% 1.03 (0.95, 1.12) 1.00 (0.92, 1.09)
Educational attainment
 High school (Ref) 1.00 1.00
 < high school 1.28** (1.10, 1.48) 1.12 (0.96, 1.30)
 Some college 1.10 (1.00, 1.21) 1.13* (1.02, 1.25)
 ≥ college 1.09 (0.99, 1.21) 1.17** (1.06, 1.30)
Employed 0.65*** (0.59, 0.71) 0.96 (0.88, 1.05)
Region of residence
 Northeast (Ref) 1.00 1.00
 Midwest 0.84** (0.74, 0.96) 0.83** (0.73, 0.94)
 South 1.04 (0.92, 1.17) 1.01 (0.90, 1.14)
 West 0.94 (0.82, 1.07) 0.92 (0.80, 1.06)
Any minor children in household 0.82*** (0.75, 0.90) 0.91* (0.83, 1.00)
Fair or poor health 2.08*** (1.84, 2.35)
Serious psychological distress 1.55*** (1.26, 1.90)
Activity limitation 2.20*** (1.94, 2.50)
Close relative needs long-term care 2.15*** (1.89, 2.43)
F 5.65* 25.33*** 47.42***
Open in a new tab

Note. CI = confidence interval; OR = odds ratio.

*P < .05; **P < .01; ***P < .001.

In model 3, after adjustment for health characteristics and having a close relative who had needed LTC, the association between LGB status and LTC expectations became nonsignificant. Fair or poor health status (OR = 2.08; 95% CI = 1.84, 2.35; P < .001), psychological distress (OR = 1.55; 95% CI = 1.26, 1.90; P < .001), activity limitations (OR = 2.20; 95% CI = 1.94, 2.50; P < .001), and having a close relative who had needed LTC (OR = 2.15; 95% CI = 1.89, 243; P < .001) were all associated with higher odds of expecting to need LTC in the future.

Regardless of whether they believed they might need LTC in the future, respondents differed according to sexual orientation in what type of care they expected to use (Figure 1). Compared with heterosexual adults, LGB adults were less likely to expect to rely on family (OR = 0.71; 95% CI = 0.52, 0.97; P < .05) and more likely to expect to use a nursing home or assisted living facility (OR = 1.53; 95% CI = 1.04, 2.27; P < .05) after control for sociodemographic and health characteristics. Odds of expecting to hire someone and expecting to use a home health care organization for LTC were similar among sexual minority group members.

FIGURE 1—

FIGURE 1—

Open in a new tab

Adjusted odds of expected use of different types of long-term care, by sexual orientation: National Health Interview Survey sample adult component, United States, 2013.

Note. LGB = lesbian, gay, or bisexual. Values are from logistic regression models adjusting for gender, age, relationship status, race/ethnicity, family income, educational attainment, employment status, region of residence, presence of minor children in the household, self-reported health status, psychological distress, presence of an activity limitation, and having a close relative who has needed long-term care.

*P < .05.

DISCUSSION

This study presents the first nationally representative data, to our knowledge, on how LTC expectations vary by sexual orientation. We found evidence that LGB adults have elevated expectations that they will need LTC in the future; however, that association is explained by sociodemographic and health characteristics (although the odds of expected need for LTC remained higher for LGB individuals, they were no longer statistically significant when sociodemographic characteristics were included in the model). Research suggests that such differences in health are closely tied to the experience of minority stress and discrimination over the life course.12–16 Our study provides further evidence that the resulting health disparities may have long-term consequences for health status and use of health care services.

Interestingly, respondents who had never been married had increased odds of expecting to need LTC in the future. LGB individuals were much more likely than heterosexual individuals to have never been married (51% vs 13%), potentially owing to limited access to legal marriage (prior to June 26, 2015, approximately half of US states did not allow same-gender couples to legally wed). In addition to being unmarried, being highly educated (college degree or beyond), having a current activity limitation, and having a close relative who had needed LTC were all associated with higher odds of expecting to need LTC in the future, and LGB individuals were more likely than their heterosexual counterparts to fall into each of these categories. Although there were also differences in expectations according to health status and psychological distress, we found no discernible differences in those characteristics by sexual orientation.

Consistent with previous studies, we found that many adults do not believe they will need LTC in the future.20 In fact, more than half of all heterosexual respondents (61%) and half of all LGB respondents believed that it was either very unlikely or unlikely that they would need LTC. This finding does not align with estimates indicating that the majority of Americans will need LTC at some point.18,19 Our results should raise concern about a general lack of planning and awareness regarding LTC. If individuals do not expect to need LTC, it is less likely that they are making arrangements, financially or otherwise, for such care and more likely that they will be unprepared should a need arise.

Regardless of whether individuals expect to need LTC in the future, most respondents in our study indicated that they would rely on family rather than on formal systems of care. This result does align with reality. Currently, family caregivers provide the bulk of LTC in the United States.44 Public health practitioners and researchers should be concerned that fewer people in our sample expected to use formal systems of care as opposed to, or in addition to, family care. Expecting to rely on unpaid family caregiving may indicate a lack of LTC planning. LGB adults in our study were significantly less likely than heterosexual adults to expect to rely on family members, perhaps because they were less likely to have children or kin caregivers on whom they could rely.2,24 This situation may place LGB older adults at increased risk for using more expensive, formal systems of care or having unmet LTC needs related to cost.28

Limitations

Our use of NHIS data resulted in some study limitations. All NHIS responses are self-reported, which can lead to recall and response bias in descriptions of health status, health behaviors, and experiences in accessing care. In addition, self-reports of LGB status may involve selection bias.45 Lesbian women, for instance, are more likely than gay men to register and report their sexual orientation status. Adults reporting sexual minority status are more likely to be highly educated,46 and nonresponse is common among members of racial and ethnic minority groups.47 Because of the documented relationship between socioeconomic disadvantage and poor health outcomes, we may have underestimated the association between sexual orientation and LTC expectations at the lower end of the socioeconomic spectrum.

Our study would have benefited from additional information that was missing in the NHIS. For example, data on transgender identity were not included in the 2013 NHIS. Transgender individuals, although a relatively small group, are often not identified in federally sponsored health surveys. Given this group’s high risk of experiencing impaired health, homelessness, substance abuse, discrimination, and violence,4 the NHIS should continue its work to incorporate transgender identity status in future surveys.48 In addition, we included bisexual individuals with respondents who self-identified as gay or lesbian, despite the bisexual population being a distinct group. We did so because of sample size constraints (slightly more than 15% of all LGB respondents in our sample self-identified as bisexual), but future studies should attempt to assess this group separately.

Furthermore, although respondents were asked whether they had a close relative who had needed LTC, they were not asked about close friends or other unrelated persons, which may have been particularly relevant for LGB individuals. However, we still found that LGB individuals were significantly more likely than heterosexual individuals to have a close relative who needed LTC. It is possible that some respondents included “chosen” family members (i.e., not legal relatives) in their response to this question. In addition, because of the cross-sectional design of the NHIS, no data are available regarding how well current expectations align with future experiences.

Conclusions

Notwithstanding the limitations just described, ours is the first study, to our knowledge, to show that LGB adults are more likely to expect LTC in the future and more likely to expect to rely on nursing homes and assisted living facilities than their heterosexual counterparts. Given that LGB individuals are less confident that they can rely on family care in the future, there is a particular need to ensure that formal systems of care are accessible and welcoming to the LGB population. Unfortunately, research has shown that LTC systems can be unwelcoming to LGB older adults,31,32,49 whether explicitly or through a lack of recognition of systematic biases. Health care workers should be trained in LGB-specific health needs, and policies prohibiting discrimination based on sexual orientation should be considered by state governments and individual nursing homes. Tailored resources need to be developed to allow LGB individuals to identify LGB-friendly providers and institutions.

Furthermore, nursing home policies should allow partnered same-gender couples the same access to shared rooms as heterosexual couples, and state Medicaid policies should extend impoverishment protections to same-gender couples. The latter step would prevent one partner from losing his or her home or savings in order for the other to qualify for Medicaid LTC benefits.

In addition, policies designed to encourage LTC planning should be targeted toward the unique needs and concerns of the LGB population. Such policies might address financial planning, individual planning for health care decision-making, and LTC insurance coverage. Public health professionals and health care providers and payers should be mindful of where these issues fit into the larger debate about legal access to same-gender marriage among LGB adults,50 given that marriage secures benefits often used in later life such as hospital visitation rights, health directives, and shared retirement plans,2 all of which are important in planning for and successfully obtaining appropriate LTC.

Acknowledgments

This study was supported by the Minnesota Population Center’s Integrated Health Interview Series project (National Institutes of Health grants R01HD046697 and R24HD041023), funded through grants from the Eunice Kennedy Shriver National Institute for Child Health and Human Development. Additional support was provided to C. Henning-Smith through a University of Minnesota interdisciplinary doctoral fellowship, to Gilbert Gonzales through a University of Minnesota doctoral dissertation fellowship, and to Tetyana P. Shippee through grant UL1TR000114 from the National Center for Advancing Translational Sciences of the National Institutes of Health.

Human Participant Protection

No protocol approval was needed for this study because the data used were obtained from publicly available, de-identified secondary sources.

References

Articles from American Journal of Public Health are provided here courtesy of American Public Health Association

RESOURCES

OSZAR »