Is Perceived Stigma in Clinical Settings Associated with Poor Health Status among New York City’s Residents of Color?

Abstract

Objectives.

Our objectives were to assess rates of perceived stigma in health care (clinical) settings reported by racially-diverse New York City residents and to examine if this perceived stigma is associated with poorer physical and mental health outcomes.

Methods.

We analyzed data from the 2016 New York City Community Health Survey. We applied bivariable and multivariable methods to assess rates of perceived stigma, and perceived stigma’s statistical relationship with health care access, physical health status, and mental health status controlling for sociodemographics and health insurance status.

Results.

Perceived stigma was associated with poorer health care access (OR=7.07,CI=[5.32–9.41]), depression (OR=3.80,CI=[2.66–5.43]), diabetes (OR=1.86,CI=[1.36–2.54]), and poor overall general health (OR=0.43,CI=[0.33–0.57]). Hispanic respondents reported the highest rate of perceived stigma among racial and ethnic minority groups (mean=0.07,CI=[0.05,0.08]).

Conclusion.

We found that perceived stigma in health care settings was a potential barrier to good health. Prior studies have illustrated that negative health outcomes are common for patients who avoid or delay care; thus, the unfortunate conclusion is that even in a diverse, heterogeneous community, stigma persists and may negatively affect well-being. Therefore, eliminating stigma in clinical settings should be a top priority for health care providers and public health professionals seeking to improve health equity.

INTRODUCTION

New York City is one of the most populous and diverse metropolitan areas globally and is the leading point of entry for legal immigrants into the United States.^1,2 Over a third of New York City’s eight million residents are foreign-born, and a significant proportion of New York City’s residents identify as racial and ethnic minorities;^2–3 thus, making it the optimal environment in which to apply intergroup contact theory, a psychosocial theory that asserts that interpersonal contact between groups that are different from one another builds tolerance and understanding within members of those groups.^4–6 Intergroup contact theory suggests that in a society where routine contact between different groups is high (e.g., racially diverse, densely populated), stigma and acts of discrimination against members of minority groups should be low due to the near-constant exposure to and acceptance of people from diverse backgrounds.^4–6 However, some scientific studies have found that people of color (also referred to as racial and ethnic minorities) continue to perceive and experience stigma in health care settings, even in large, diverse metropoles, and this perceived stigma has serious, detrimental effects on personal well-being.^7–12 Considering these countervailing perspectives, we analyzed data from the 2016 Community Health Survey to calculate rates of clinic-based perceived stigma reported by New York City residents. We then examined associations between perceived stigma and negative mental and physical health outcomes while controlling for sociodemographic characteristics and health insurance status. We hypothesized that perceived stigma in health care settings would be low, and perceived stigma would be associated with poorer health status.

Stigma is a damaging process through which a group of individuals is labeled as socially devalued due to attributes or behaviors that are societally believed to be “deeply discrediting.”¹³ Because stigma is a social occurrence, it requires a group that has social authority (typically a majority population) as compared to the “other,” (usually a minority population); stigma is created and perpetuated by structure and interpersonal processes.¹⁴ There are three accepted domains of stigma: perceived stigma, enacted stigma, and internalized stigma.^15–17 In the case of people living with HIV, perceived and internalized stigmas’ negative health effects include non-optimal medication adherence, lower visit adherence, higher rates of depressive symptoms, overall worse quality of life, and poorer health status.^8,18–20 Studies among a range of minority groups suggest that exposure to any of these three types of stigma is associated with higher rates of poor mental health.^8,17,21,22 Since stigma reduces the social standing of members of a stigmatized population, in the case of our study, people of color, those belonging to a stigmatized group stop being regarded as “normal,” effectively increasing their likelihood of experiencing social rejection (across various environments, including health care settings).¹³ Thus, perceived stigma in clinical settings is likely to act as a barrier for New York City’s racial and ethnic minority residents to seeking necessary physical and mental health care.

Of all the environments that people access, it is in clinical health care settings where individuals are the most vulnerable due to their role transition from person to patient.²³ Therefore, health care settings (e.g., clinics, hospitals) must be stigma-free, particularly if health care stakeholders seek to provide optimal personalized health care and medical guidance that patients trust and to which they will adhere.^24–26 Prior studies have shown that although health care settings should be stigma-free, this is not always the case. Global health research has shown that sexual and gender minorities routinely encounter stigma in clinical settings;^17,27 related studies have found that multiple forms of stigma exist across various settings toward an array of minority groups (e.g., racial, gender, etc.) in the United States.^{12,21,28–29} Fewer research studies make the subsequent connection between the patients’ perceptions of stigmatizing attitudes in their clinic to their own personal physical and mental health. Thus, a novel feature of our study is that not only do we examine potential associations of perceived stigma with physical and mental health outcomes, but we do so in a city with immense cultural diversity, ensuring that most of the population, both patients and clinical providers, interact with people who are physically and socially different than themselves routinely.

Although we hypothesized that a relationship exists between perceived stigma in health care settings and physical and mental health statuses, other personal characteristics, such as household income, health insurance coverage, age, and English proficiency (applicable to non-native speakers) unquestionably influence individual and population health. English proficiency is uniquely relevant to this study since we examine the health of a population that includes many foreign-born respondents. Limited English proficiency, in and of itself, may create uncomfortable interactions between patients and health care providers that could lead to stigmatizing acts (both intentional and unintentional) from the provider toward the patient. Considering prior population health and social science research findings, we examined the relationship between perceived stigma and health in New York City.

METHODS

Using data from the 2016 New York City Community Health Survey (CHS), we conducted frequencies and bivariable and multivariable (logistic regressions) analyses of respondents’ perceptions of stigma in health care settings, health care access, and health status. For multivariable models, we performed tests of multicollinearity and found the mean variation inflation factors to be less than 2 in all cases.

Data

The New York City Department of Health and Mental Hygiene (DOHMH) conducts the Community Health Survey (CHS) annually to gather information on the health status, health care use, and health behaviors of non-institutionalized adults 18 years of age and older in the five boroughs of New York City. DOHMH surveys individuals by phone, both cellular and landline, in several languages, including English, Spanish, Russian, and Chinese. Interviewers randomly select one adult from each household to interview. Of potential respondents who were contacted and deemed eligible for the survey, 85.3% participated. All data collected were self-reported.

DOHMH collects CHS data from a stratified random sample of United Hospital Fund (UHF) designated neighborhoods in New York City.³⁰ The UHF combines several adjacent zip codes with similar characteristics into forty-two designated neighborhoods, though due to the small size of some neighborhoods, CHS collapses this number into thirty-four.³⁰ The CHS includes individual observation weights such that results can be generalized the population of New York City. Weights reflect the probability of an adult’s selection from a household, the type of phone in a household (cellular only, landline only, or both), and the size and demographic composition of each UHF neighborhood. The unweighted sample size for this study was n=9,389.

Measures

The primary independent variable of interest was perceived stigma in health care, which was assessed with the following survey question: “Thinking of your experiences trying to get health care treatment in the past 12 months, have you felt you were hassled, made to feel inferior, or discriminated against for any reason?” Responses were coded as Yes (1), No (2), Did not seek healthcare treatment in past 12 months (3), Don’t know/not sure (missing), Refused (missing). In our statistical analysis, the variable was coded as a binary variable, as the missing values were present in the original data. The CHS instruments were adapted from the Centers for Disease Control and Prevention’s BRFSS and the National Health Interview Survey.^31,32 The perceived stigma question is more general than the similar question asked in some rounds of the Behavioral Risk Factor Surveillance System surveys.³³

The outcome measures consisted of dummy variables indicating health care access, mental health, and physical health. The variable “Did not get needed care” was assessed with the following survey item: “Was there a time in the past 12 months when you needed medical care but did NOT get it? Medical care included doctor visits, tests, procedures, prescription medication and hospitalizations. (Yes, No).” Next, to assess if the respondent had to seek medical care from an informal source, we used the question “When you are sick or need advice about your health, to which of the following places do you usually go? (1= A private doctor, 2= Non-retail clinic, 3= Urgent Care Center, 4= Hospital ED, 5= Retail clinic, 6=Alternative health care provider such as traditional healer or herbalist), 7= Family/friend/self/internet resources, 8= Other, 9= No usual place).” Responses were combined into 2 categories: 0 for options 1–5, 1 for 6–9 respectively; this created the variable “Medical Advice from Informal Sources.” Mental health dummy variables indicated if an individual reported being depressed within the past two weeks, based on their responses to the eight-item Patient Health Questionnaire (PHQ8), a validated scale of depression. Additionally, among depressed individuals, if depression has made it somewhat, very, or extremely difficult to do work, take care of things at home, or get along with other people, the variable “depression impacts daily life” equaled one and zero otherwise. Physical health dummy variables indicated whether an individual reported that his or her health was good, very good, or excellent (bad, very bad being the referent group); whether a health professional had ever told the respondent that they had hypertension; and if a health professional had ever told the respondent that they had diabetes (excluding cases of gestational diabetes).

In our analyses, we controlled for several sociodemographic characteristics. Health insurance status was reported in two steps. In the first step, the respondent answered a yes/no question on health insurance coverage. For individuals reporting any form of coverage, further questions were asked on the type of coverage. We combined these responses in a categorical variable, where the reference category was private insurance since it is the largest group. Race and ethnicity dummy variables indicated if the respondent was black (non-Hispanic), Hispanic, Asian or Pacific Islander (non-Hispanic), or other non-Hispanic, with a reference category of white. Age dummy variables indicated the following categories: 18–24 (reference category); 25–44; 45–64; and 65 or more years. Earnings dummy variables indicate the federal poverty level (FPL) category of annual household earnings: less than 100% (reference category); 100 to less than 200%; 200 to less than 400%; 400 to less than 600%; and 600% or more. Other demographic dummy variables indicated whether the individual was born in the United States, identified as male, was married (including cohabiting), and primarily spoke a language other than English at home. Throughout our dataset, the responses of “don’t know” and “refused” were coded as missing by the DOHMH.

Analysis

For the descriptive statistics, we report weighted proportions. We reported means of outcome variables, both for the full sample and across groups who did or did not report perceived stigma. For the multivariable analyses, we reported the odds ratios obtained from logistic regressions of perceived stigma on each outcome, with 95% confidence intervals in brackets below each estimate. A single asterisk (*) indicated a p-value of less than 0.1; a double asterisk (**) indicated a p-value of less than 0.5; and a triple asterisk (***) indicated a p-value of less than 0.01. All descriptive statistics and regression estimates were weighted to reflect the complex survey design. We analyzed the results in Stata version 14 using its suite of svy commands and weights from the NYC-CHS survey.

Ethical Approval

Ethical approval was provided by the City University of New York (CUNY) Institutional Review Board (IRB, #2018–0473).

RESULTS

Descriptive statistics

We reported descriptive statistics for control variables in Table 1. Thirty-five percent of the sample was non-Hispanic white (henceforth referred to as white), 22% non-Hispanic black (subsequently referred to as black), 27% Hispanic, 13% Asian or Pacific Islander, and 2% other. Compared to a representative survey of New York City residents from the 2017 American Community Survey (ACS),^34, our sample had a smaller proportion of white individuals (35% compared to 43%). Less than half the sample (46%) was male, 42% were married, 60% were employed, and 34% were college graduates. Over half of the sample (52%) was born in the United States, compared to 63% citywide. Thirty-five percent lived in households where the primary language was not English. The majority (84%) had a primary care doctor. Thirteen percent of the sample that was 18–24 years, 40% were 25–44 years, and 32% were 45–64 years. About half of the sample (47%) had private insurance, compared to 16% Medicare, 24% Medicaid, 3% other insurance, and 11% uninsured. About a quarter of the households in the sample (26%) fell below the federal poverty line (FPL), compared to 20% citywide.³⁴ Twenty-two percent of the sample fell between 100 and 199% FPL, 17% between 200 and 399% FPL, 16% between 400 and 599% FPL, and 19% 600% or more FPL.

Table 1:

Descriptive statistics of the 2016 New York City Community Health Survey (CHS, n=9389)

Race
White	35%
Black	22%
Hispanic	27%
Asian/Pacific Islander	13%
Other Non-Hispanic	2%
Demographics
Male	46%
Married	42%
Employed	60%
College Graduate	34%
U.S.-born	52%
Non-English at Home	35%
Has a Primary Care Doctor	84%
Age
18–24	13%
25–44	40%
45–64	32%
65+	15%
Insurance
Private	47%
Medicare	16%
Medicaid	24%
Other	3%
Uninsured	11%
Poverty (% Federal Poverty Level, FPL)
<100% FPL	26%
100 – <200% FPL	22%
200 – <400% FPL	17%
400 – <600% FPL	16%
>600% FPL	19%

Note: All percentages are weighted using the survey weights provided by the NYC Community Health Survey

Source: Authors’ calculations based on 2016 NYC CHS data

Hispanic and black respondents reported the highest rates of perceived stigma (6.7% [0.05,0.08] and 6.2% [0.05,0.08], respectively). Asian and Pacific Islander respondents reported the lowest rate (3.6% [0.02,0.05]). Nearly 6% (5.7% [0.04,0.07]) of white individuals reported perceiving stigma in their health care settings. Table 2 contains descriptive statistics of the predictor variables, classified by perceived stigma in health care settings. All differences between these groups are statistically significant at the 95% level. Respondents who perceived stigma reported higher rates of not receiving needed care (63% compared to 17%) and of regularly receiving advice from informal sources (8% compared to 4%). These stigmatized respondents reported higher rates of depression (26% compared to 8%), and among the depressed, higher rates of depression interfering in daily activities (88% compared to 82%). Stigmatized respondents reported lower rates of good health (35% compared to 43%) and higher rates of hypertension (50% compared to 48%).

Table 2:

Percentages of individuals reporting each health outcome, by perceived stigma: 2016 NYC Community Health Survey

	Full Sample	No Stigma	Perceived Stigma
Did not get needed medical care	24.0%	17.0%	63.0%
Seek health care from an informal source	4.0%	4.0%	8.0%
Current depression	9.0%	8.0%	26.0%
Depression negatively impacts daily life (among depressed)	83.0%	82.0%	88.0%
General health is excellent, very good, or good	42.0%	43.0%	35.0%
Hypertension	49.0%	48.0%	50.0%
Diabetes	19.0%	20.0%	15.0%

Note: All percentages are weighted. All differences are statistically significant at the 95 percent level except for the variable “Depression negatively impacts daily life (among depressed).

Source: Authors’ calculations based on 2016 NYC CHS data

Multivariable Analyses

In Table 3, we reported odds ratios from logistic regression models of perceived stigma on health care access indicators. Columns 1 and 3 presented results of bivariable regressions; columns 2 and 4 include all control variables. Our main variable of interest, perceived stigma, was associated with an increase in the odds of not accessing needed medical care (Odds Ratio, OR=8.03, 95% Confidence Interval, CI=[6.08–10.59] and Adjusted Odds Ratio, AOR=7.07, CI=[5.32–9.41] respectively). Perceived stigma was also associated with an increase in the odds of receiving medical care and advice from an informal source (OR=2.92, CI=[1.63–5.24] and AOR=2.26, CI=[1.29–3.95] respectively). The adjusted odds ratios were obtained after controlling for respondents’ race and ethnicity, sex, age, income, education, type of insurance (or lack of it), nativity (where a respondent was born), and primary language spoken at home.

Table 3:

Unadjusted and adjusted odds ratios from logistic regressions of perceived stigma on health care access outcomes: 2016 NYC Community Health Survey

	Did not get needed care		Medical advice from informal source
	(1)	(2)	(3)	(4)
	OR	AOR	OR	AOR
Perceived Stigma	8.03***	7.07***	2.92***	2.26***
	[6.08 – 10.59]	[5.32 – 9.41]	[1.63 – 5.24]	[1.29 – 3.95]
Insurance
Ref: Private
Medicare		1.69***		0.73
		[1.15 – 2.48]		[0.39 – 1.37]
Medicaid		1.84***		1.16
		[1.37 – 2.47]		[0.65 – 2.04]
Other		1.77		1.73
		[0.87 – 3.62]		[0.79 – 3.76]
Uninsured		2.00***		3.80***
		[1.40 – 2.87]		[2.19 – 6.60]
Demographics
U.S.-born		0.97		0.74
		[0.73 – 1.29]		[0.49 – 1.12]
Male		1.12		1.72***
		[0.92 – 1.38]		[1.20 – 2.48]
Married		0.76**		0.95
		[0.60 – 0.95]		[0.65 – 1.38]
Non-English at Home		0.76		1.38
		[0.53 – 1.09]		[0.85 – 2.24]
Race
Ref: White Non-Hispanic
Black Non-Hispanic		0.90		0.88
		[0.65 – 1.23]		[0.51 – 1.55]
Hispanic		1.01		0.99
		[0.74 – 1.39]		[0.61 – 1.60]
Asian/PI Non-Hispanic		0.81		0.57
		[0.53 – 1.23]		[0.27 – 1.19]
Other		1.30		1.06
		[0.67 – 2.52]		[0.44 – 2.56]
Age
Ref: 18–24 yrs
25–44 yrs		1.47**		0.97
		[1.01 – 2.14]		[0.56 – 1.67]
45–64 yrs		1.18		0.54**
		[0.81 – 1.72]		[0.30 – 0.96]
65+ yrs		0.72		0.66
		[0.45 – 1.15]		[0.34 – 1.28]
Poverty
Ref.<100% FPL
100 – <200% FPL		0.98		0.82
		[0.73 – 1.31]		[0.53 – 1.27]
200 – <400% FPL		1.16		0.87
		[0.83 – 1.63]		[0.49 – 1.53]
400 – <600% FPL		0.94		0.99
		[0.64 – 1.37]		[0.55 – 1.78]
>600% FPL		0.63**		1.07
		[0.41 – 0.97]		[0.55 – 2.09]
Constant	0.09***	0.07***	0.04***	0.03***
	[0.08 – 0.10]	[0.04 – 0.13]	[0.03 – 0.04]	[0.01 – 0.07]
Observations	9,437	9,437	9,386	9,386

Notes: Perceived Stigma =1 if the respondent answers YES to the question “Thinking of your experiences trying to get health care treatment in the past 12 months, have you felt you were hassled, made to feel inferior, or discriminated against for any reason?” All models are estimates using the survey weights reflecting the complex survey designs used by NYC Community Health Survey. 95 percent confidence intervals in brackets below odds ratios (OR) and adjusted odds ratios (AOR).

FPL: Federal Poverty Level;

^***p<0.01,

^**p<0.05,

^*p<0.1

Source: Authors’ calculations based on 2016 NYC CHS data

Table 4 reported odds ratios from logistic regressions of perceived stigma on mental health indicators. As in Table 3, columns 1 and 3 presented results of bivariable regressions, while columns 2 and 4 include all control variables. Perceived stigma was associated with an increase in the odds of having a diagnosis of depression (AOR=3.80, CI=[2.66–5.43]). Additionally, being married was associated with a 34% decrease in the odds of depression. Higher income was associated with lower rates of depression (p<0.01 for each estimate). Among those with a depression diagnosis, perceived stigma was associated with an unadjusted 75% increase and an adjusted 56% increase in the odds of depression, making daily life difficult.

Table 4:

Unadjusted and adjusted odds ratios from logistic regressions of perceived stigma on mental health outcomes: 2016 NYC Community Health Survey

	Depressed		Depression impacts daily life
	(1)	(2)	(3)	(4)
	OR	AOR	OR	AOR
Perceived Stigma	3.91***	3.80***	1.75	1.56
	[2.81 – 5.46]	[2.66 – 5.43]	[0.89 – 3.46]	[0.74 – 3.30]
Insurance
Ref: Private
Medicare		1.86***		0.67
		[1.19 – 2.90]		[0.31 – 1.47]
Medicaid		1.79***		1.29
		[1.26 – 2.54]		[0.62 – 2.71]
Other		1.05		0.69
		[0.55 – 2.01]		[0.20 – 2.41]
Uninsured		1.17		1.39
		[0.71 – 1.93]		[0.52 – 3.75]
Demographics
U.S.-born		1.28		1.94**
		[0.95 – 1.71]		[1.01 – 3.76]
Male		0.89		0.96
		[0.71 – 1.12]		[0.57 – 1.63]
Married		0.63***		0.50**
		[0.49 – 0.82]		[0.29 – 0.87]
Non-English at Home		0.84		1.14
		[0.58 – 1.21]		[0.54 – 2.42]
Race
Ref: White Non-Hispanic
Black Non-Hispanic		0.65**		0.69
		[0.46 – 0.94]		[0.31 – 1.56]
Hispanic		0.99		0.65
		[0.70 – 1.42]		[0.32 – 1.31]
Asian/PI Non-Hispanic		0.53**		1.13
		[0.31 – 0.91]		[0.36 – 3.61]
Other		1.15		2.08
		[0.57 – 2.33]		[0.28 – 15.51]
Age
Ref: 18–24 yrs
25–44 yrs		1.16		3.44**
		[0.76 – 1.75]		[1.32 – 8.98]
45–64 yrs		1.75***		2.59**
		[1.16 – 2.64]		[1.04 – 6.43]
65+ yrs		1.09		1.85
		[0.64 – 1.84]		[0.65 – 5.30]
Poverty
Ref.<100% FPL
100 – <200% FPL		0.61***		0.60
		[0.44 – 0.83]		[0.32 – 1.11]
200 – <400% FPL		0.49***		0.56
		[0.34 – 0.71]		[0.27 – 1.18]
400 – <600% FPL		0.28***		1.46
		[0.16 – 0.46]		[0.60 – 3.56]
>600% FPL		0.17***		0.46
		[0.09 – 0.29]		[0.14 – 1.57]
Constant	0.09***	0.13***	4.38***	2.62
	[0.08 – 0.10]	[0.06 – 0.25]	[3.32 – 5.78]	[0.72 – 9.49]
Observations	8,872	8,872	9,369	9,369

Notes: Perceived Stigma =1 if the respondent answers YES to the question “Thinking of your experiences trying to get health care treatment in the past 12 months, have you felt you were hassled, made to feel inferior, or discriminated against for any reason?” All models are estimates using the survey weights reflecting the complex survey designs used by NYC Community Health Survey. 95 percent confidence intervals in brackets below odds ratios (OR) and adjusted odds ratios (AOR).

FPL: Federal Poverty Level;

^***p<0.01,

^**p<0.05,

^*p<0.1

Source: Authors’ calculations based on 2016 NYC CHS data

Table 5 reported results from logistic regressions of perceived stigma on physical health indicators. Association of perceived stigma with physical health was qualitatively similar to those of mental health. Perceived stigma was associated with a decrease in the odds of being in a good, very good, or excellent health (OR=0.43, CI=[0.33–0.57] and AOR=0.36, CI=[0.26–0.50] respectively). It was also associated with an increase in the odds of having been diagnosed with hypertension (OR=1.52, CI=[1.16–1.99] and AOR=0.36, CI=[0.26–0.50] respectively). Finally, columns 5 and 6 show that perceived stigma was associated with an increase in the odds of having been diagnosed with diabetes (OR=1.44, CI=[1.01–2.04] and AOR=0.36, CI=[0.26–0.50] respectively).

Table 5:

Unadjusted and adjusted odds ratios from logistic regressions of perceived stigma on physical health outcomes: 2006 NYC Community Health Survey

	Good Overall Health		Hypertension		Diabetes
	(1)	(2)	(3)	(4)	(5)	(6)
	OR	AOR	OR	AOR	OR	AOR
Perceived Stigma	0.43***	0.36***	1.52***	1.86***	1.44**	1.68***
	[0.33 – 0.57]	[0.26 – 0.50]	[1.16 – 1.99]	[1.36 – 2.54]	[1.01 – 2.04]	[1.16 – 2.45]
Insurance
Ref: Private
Medicare		0.51***		1.68***		1.36**
		[0.39 – 0.66]		[1.33 – 2.13]		[1.03 – 1.80]
Medicaid		0.69***		1.12		1.03
		[0.54 – 0.87]		[0.90 – 1.40]		[0.78 – 1.35]
Other		0.76		0.95		0.66
		[0.48 – 1.22]		[0.64 – 1.40]		[0.38 – 1.15]
Uninsured		0.74*		0.80		0.70
		[0.53 – 1.03]		[0.58 – 1.11]		[0.43 – 1.15]
Demographics
U.S.-born		0.88		1.26**		0.96
		[0.71 – 1.10]		[1.05 – 1.51]		[0.75 – 1.23]
Male		1.18**		0.96		1.13
		[1.01 – 1.39]		[0.83 – 1.12]		[0.94 – 1.36]
Married		1.02		0.87*		1.04
		[0.86 – 1.22]		[0.75 – 1.02]		[0.85 – 1.27]
Non-English at Home		0.69***		1.09		0.82
		[0.54 – 0.88]		[0.87 – 1.38]		[0.61 – 1.10]
Race
Ref: White Non-Hispanic
Black Non-Hispanic		0.96		2.12***		1.96***
		[0.74 – 1.24]		[1.70 – 2.64]		[1.45 – 2.66]
Hispanic		0.83		1.68***		2.29***
		[0.65 – 1.06]		[1.33 – 2.14]		[1.72 – 3.07]
Asian/PI Non-Hispanic		0.56***		0.88		1.43*
		[0.42 – 0.75]		[0.66 – 1.18]		[0.96 – 2.13]
Other		0.58*		1.33		1.69*
		[0.33 – 1.03]		[0.79 – 2.23]		[0.98 – 2.91]
Age
Ref: 18–24 yrs
25–44 yrs		0.43***		2.33***		3.62**
		[0.29 – 0.65]		[1.53 – 3.56]		[1.31 – 10.06]
45–64 yrs		0.15***		10.59***		20.94***
		[0.10 – 0.23]		[7.04 – 15.94]		[7.75 – 56.59]
65+ yrs		0.12***		20.57***		37.59***
		[0.08 – 0.19]		[13.48 – 31.38]		[13.75 – 102.74]
Poverty
Ref.<100% FPL
100 – <200% FPL		1.27**		0.86		0.76**
		[1.03 – 1.57]		[0.69 – 1.08]		[0.60 – 0.96]
200 – <400% FPL		2.03***		0.85		0.68**
		[1.55 – 2.65]		[0.66 – 1.09]		[0.50 – 0.92]
400 – <600% FPL		2.80***		0.66***		0.57***
		[2.05 – 3.83]		[0.51 – 0.85]		[0.41 – 0.79]
>600% FPL		5.64***		0.59***		0.32***
		[3.86 – 8.23]		[0.44 – 0.79]		[0.21 – 0.49]
Constant	3.79***	15.81***	0.38***	0.05***	0.12***	0.01***
	[3.52 – 4.08]	[8.88 – 28.13]	[0.35 – 0.40]	[0.03 – 0.08]	[0.11 – 0.13]	[0.00 – 0.02]
Observations	9,359	9,359	9,411	9,411	9,415	9,415

Notes: Perceived Stigma =1 if the respondent answers YES to the question “Thinking of your experiences trying to get health care treatment in the past 12 months, have you felt you were hassled, made to feel inferior, or discriminated against for any reason?” All models are estimates using the survey weights reflecting the complex survey designs used by NYC Community Health Survey. 95 percent confidence intervals in brackets below odds ratios (OR) and adjusted odds ratios (AOR).

FPL: Federal Poverty Level;

^***p<0.01,

^**p<0.05,

^*p<0.1

Source: Authors’ calculations based on 2016 NYC CHS data

DISCUSSION

Consistent with expectations, perceived stigma was associated with lower rates of health care access, higher rates of depression, increased severity of depressive symptoms, and worse physical health, even after adjusting for individual confounders. We found mixed support for the application of Intergroup Contact Theory in New York City; Intergroup Contact Theory suggests that stigma should be lower, if not non-existent, in heterogeneous societies. Although the rates of perceived stigma were lower than found in sample studies conducted elsewhere and in studies conducted with other minority groups,^8,10,21,27 perceived stigma was still reported by our New York City respondents, leading to questions as to why this is the case. Was there a sociocultural shift in 2016 (when data was collected) that could have affected stigmatizing attitudes (both in the clinic and as perceived by the patient)? Is there potentially a baseline level of stigma that can be expected? If so, is that baseline near 6% as found in our black and white respondents or is it closer to 4%, the rate reported by Asian respondents in our sample? Considering the ongoing dialogue about generational differences, are there significant generational differences in those who report perceiving stigma? If so, this could imply the need for health care settings to be culturally open to not only racial and ethnic differences, but also to those related to being younger and to identifying as a sexual and gender minority.

Together, our findings suggest that perceived stigma may be a barrier to good health. Stigma was especially pronounced among racial and ethnic minority groups. Hispanic respondents reported the highest rates of perceived stigma, potentially reflecting the combined effect of ethnicity and language (intersectionality). Even some white respondents reported experiencing perceived stigma in health care settings. Although initially surprising, this finding may be explained by methodological/data gaps. Specifically, white respondents who identified as a sexual or gender minority could perceive stigma due to their identity or orientation. Alternatively, white respondents who were immigrants could perceive stigma due to their immigrant status or if their spoken English was accented. Both potential explanations point to the impact of intersectionality or intersectional stigma, which is a form of stigma affecting populations that holding multiple stigmatizing characteristics (e.g., black men who have sex with men, Hispanic immigrants, etc.).

Ultimately, the relationship between increased heterogeneity (diversity) and stigma (perceived, enacted, and internalized) is nuanced; our study is a stepping stone to better understanding the mechanisms that lead to the creation and perpetuation of stigma. Though we found limited previous research that focused exclusively on New York City, there was some preliminary evidence against the assertions of Intergroup Contact Theory suggesting that in more hetergenous, diverse communities, higher perceptions of racism existed.^35,36 In communities that undergo diversification, members of majority groups may newly encounter the experience of being “othered” and therefore potentially feel stigmatized. For example, if a white New York City resident obtains health care at a clinic that is staffed by predominantly providers of color, this resident may perceive stigma within that health care experience, being that the patient could be considered a minority in that very specific context.

Limitations

Findings from this study should be understood within the context of its limitations. The CHS is a cross-sectional survey, and therefore, we cannot determine the direction of causality. For example, some confounding variables, such as provider race and provider-patient language concordance are not available in the dataset, and may affect both perceived stigma and health outcomes.^37,38 These types of associations may reflect institutional rather than individual racism, but are still relevant to our inquiry of study. While we controlled for a host of individual characteristics available in CHS, we did not have information on factors such as neighborhood demographics and healthcare infrastructure such as distance from providers and types of providers available.

CONCLUSION

Our innovative study sheds new light on the association between perceived stigma and health status; we found that perceived stigma was associated with poorer physical and mental health. Perceived stigma in health care settings may act as a barrier to accessing health care services, and perceived stigma may significantly harm those being stigmatized. Even in New York City, a metropolis with immense population diversity, perceived stigma persisted. It is likely that rates of perceived stigma are higher in other less diverse, less population dense metropoles, highlighting an urgent need to address this matter nationally. If we use New York City as a case study, and perceived stigma is potentially a deleterious force even in an immensely heterogeneous, diverse society, one can hypothesize that perceived (and enacted and internalized) stigma may be higher in less diverse, less population-dense cities, and this stigma could subsequently marginalize people of color (and sexual and gender minorities) who live in those communities. Local populations should be encouraged to report egregious acts of stigma and discrimination, and local chapters of providers groups should make stigma reduction a top priority to improve public health.