Associations of depression with hypertension and citizenship among U.S. adults: A cross-sectional study of the interactions of hypertension and citizenship

Highlights • Depression is common among patients with hypertension.• Disparities exist in hypertension and depression by citizenship status in the U.S.• U.S. citizens reported a higher prevalence of depression than non-citizens.• Non-U.S. citizens with hypertension had higher odds of depression compared to U.S. citizens without hypertension.


Introduction
Hypertension is prevalent in the United States (U.S.) with an estimated 47 % of adults having this condition (Centers for Disease Control and Prevention, 2021a, b).It increases the risk for heart disease and stroke, the two leading causes of death in the United States (Kochanek et al., 2019;Murphy et al., 2021).Hypertension-related cardiovascular death has significantly increased over time (Nambiar et al., 2020).In 2019, more than half a million deaths in the U.S. had hypertension as a primary or contributing cause of death (Centers for Disease Control and Prevention, 2021b).
Depression is a common comorbidity among patients with hypertension (Marazziti et al., 2014).The prevalence of depression is higher in hypertensive patients than in the general population (Kessing et al., 2020).A systematic review and meta-analysis using 41 studies found that approximately one-third of patients with hypertension also present with depression (Li et al., 2015).Existing evidence indicates that both hypertension and depression share common biological pathways including activation of the renin-angiotensin-aldosterone system (DeJean et al., 2013;Hafner et al., 2013;Rubio-Guerra et al., 2013).While previous studies have found the use of antidepressants to increase blood pressure through multiple effects on various pathways and systems in the body (Breeden et al., 2018;Calvi et al., 2021), antihypertensive medications can also affect the risk of depression (Kessing et al., 2020;Li et al., 2021).A recent nationwide study was conducted using the Danish population-based register to investigate whether commonly used antihypertensive medications were associated with depression (Kessing et al., 2020).Contrary to previous findings, the authors reported that no antihypertensive drug was associated with an increased risk of depression but rather the continued use of angiotensin agents, calcium antagonists, and β-blockers significantly reduce the rates of depression (Kessing et al., 2020).
In general, depression varied by sociodemographic characteristics (e. g., age, sex), with individuals with lower socioeconomic status (e.g., females, younger adults, unemployed) having higher burdens of depression (Gabriel et al., 2021;Mann et al., 2022;Villarroel and Terlizzi, 2020).Several other studies have examined the association between hypertension and depression and have reported mixed findings (Ginty et al., 2013;Grimsrud et al., 2009;Ho et al., 2015;Maatouk et al., 2016;Rubio-Guerra et al., 2013;Wang et al., 2021;Wiehe et al., 2006).These studies, however, did not delineate differences found in the current literature among underserved populations including minorities and immigrants (Commodore-Mensah et al., 2021;Gabriel et al., 2021;Guadamuz et al., 2020).For instance, citizenship status has been found to contribute to disparities in hypertension and other cardiovascular conditions (Guadamuz et al., 2020).A study on the association between immigration status and anxiety, depression, and use of anxiolytic and antidepressant medications in the Hispanic Community also reported that undocumented Hispanic non-citizens in the U.S. were less likely to use medication for depression compared to citizens and documented non-citizens (Ross et al., 2019).
The U.S. immigrant population has been increasing rapidly and is expected to be a major driver of population growth (Guadamuz et al., 2020;Jordan and Gebeloff, 2022).New immigrants are known to have lower incidence of chronic health conditions, a phenomenon described as the healthy immigrant paradox (Bustamante et al., 2021;Stanek et al., 2020).Nonetheless, disparities in healthcare resources, social disadvantages, neighborhood conditions, and low socioeconomic status associated with excess burden of cardiovascular diseases have also been well-documented among immigrants (Chang, 2019;de Mestral and Stringhini, 2017;Hamad et al., 2020).This poses a major health risk of further increase in the prevalence of hypertension and other chronic conditions in this population.For example, immigrants are more likely to experience limited access to care including inequitable access/treatment and low adherence to medications (Cho et al., 2020;Guadamuz et al., 2021;Hacker et al., 2015).Yet, no study to date has investigated citizenship and hypertension status disparities in depression among U.S. adults using nationally representative data.
This study aims to: (1) estimate the prevalence of hypertension and depression among U.S. adults, (2) determine the associations of depression with hypertension, citizenship status, and the interaction between hypertension and citizenship status, and (3) sociodemographic and other factors among U.S. adults.

Study participants and design
Data from the 2015-2018 National Health Interview Survey (NHIS) (Blewett et al., 2021) were analyzed.The NHIS is a cross-sectional household interview survey that targets the civilian noninstitutionalized population of the U.S. Thus, weighted estimates cover only the civilian noninstitutionalized household population.Individuals in longterm care institutions, correctional facilities, and U.S. nationals living in foreign countries were excluded.Active-duty Armed Forces personnel are also excluded from the survey, unless at least one other family member is a civilian eligible for the survey but they were not weighted (National Center for Health Statistics, 2019c).
Analyses were restricted to adults (ages 18 and older) who received questions about depression.The analyses were weighted using the NHIS sampling weights.The sampling weights are derived from person-level weights indicating the number of population units that a sampled unit represents.The person-level weights are adjusted for nonresponse and ratio-adjusted to create final sampling weights, which are further adjusted according to a quarterly post stratification by age/sex/race/ ethnicity classes based on population estimates produced by the U.S. Census Bureau (Parsons et al., 2014).Also, oversampling was applied to minority household populations.There were 72,159 participants in the 2015-2018 datasets.However, we performed a complete case analysis on 63,985 individuals.The remaining 8,174 participants were excluded due to incomplete or missing data.In this study, depression was the outcome of focus, with hypertension and citizen status as primary independent variables or exposures of interest, and sociodemographic factors as secondary independent variables.

Outcome
To assess the outcome, depression, of this study, respondents were categorized as depressed if they reported feeling depressed "daily, weekly, monthly or a few times per year or more" or if they reported "yes" that they take medication for depression.Otherwise, respondents were categorized as not depressed.To address the issue of under identification of depression, especially among non-citizens, we included those who use medication for depression.For instance, it has been reported that immigrants and other racial/ethnic minorities may not report being depressed even if they have agreed to treatment due to the stigma attached to this condition (Chen et al., 2016).Hence, it was essential to include individuals taking medication for depression to minimize sampling error.

Primary independent variables
Independent variables in this study include hypertension and citizenship status.Hypertension was categorized into two: having hypertension or not having hypertension as the reference [Ref] category.Respondents were categorized as having hypertension if they reported that "they had hypertension or high blood pressure in the past 12 months" or were "taking any medication prescribed by a doctor for high blood pressure at the time of the interview."Individuals' citizenship status was classified as being a U.S. citizen [Ref] or non-U.S.citizen.
Respondents were asked whether they are U.S. citizens or non-citizens, and they responded yes or no.

Statistical analyses
We first assessed the prevalence of depression and hypertension along with descriptive characteristics of sociodemographic and other factors using frequencies and percentages.Then two multivariable logistic regression models were used to examine (1) the associations of depression with hypertension status, citizenship status, and sociodemographic factors (2) interaction of hypertension status and citizenship status, adjusting for the sociodemographic factors.Additionally, predicted marginal proportions for depression by citizenship status and hypertension were also calculated.Analyses were weighted, and Taylor series linearization methods were used to account for the stratified, multistage, cluster sample design of the NHIS (National Center for Health Statistics, 2019a).Results are reported based on odds ratio (OR) along with 95 % confidence interval (CI) and p-values at 0.05 statistical level of significance.Analyses were conducted using SUDAAN 11.0.3.

Multivariable logistic regression
Table 2 presents the results of the multivariable logistic regression, which depicts the associations between depression, U.S. citizenship status, hypertension, intersection of citizenship and hypertension status, and other factors.Both models 1 (without interaction) and 2 (with interaction) indicate that individuals with hypertension had slightly For model 2 (with interaction model), there was a statistically significant interaction between hypertension and citizenship status (Wald F = 3.98, p = 0.009).Higher odds of depression was observed among non-US citizens with hypertension relative to U.S. citizens without hypertension (OR = 1.46; 95 % CI = 1.15, 1.86).Non-U.S. citizens without hypertension had lower odds of depression while U.S. citizens with hypertension had higher odds of depression compared to U.S. citizens without hypertension, although not statistically significant.

Predicted marginal results
Fig. 1, the predicted marginal proportions for hypertension and depression by citizenship status, highlights the interaction effects of citizenship status and hypertension in the logistic regression model.Overall, non-citizens with hypertension had the highest predicted marginal for depression (47.3 %) followed by U.S. citizens with hypertension (39.4 %), citizens without hypertension (38.9 %) and non-citizens without hypertension (37.2 %).

Discussion
Using nationally representative data from NHIS, this study demonstrates dynamic differences in the presentation of hypertension and depression by citizenship status.To the best of our knowledge, this is the first study to examine the prevalence of hypertension and depression in the general U.S. adult population, and to highlight differences based on citizenship status.Our finding on the prevalence of depression among hypertensive U.S. adults is similar to what has been reported in a systematic review and meta-analysis (Li et al., 2015).Of note, both the interaction and predicted marginal analyses consistently indicate that non-U.S.citizens with hypertension had an elevated likelihood of depression compared to U.S. citizens without hypertension although non-citizens had a lower prevalence of depression.Further, the fact that hypertension and citizenship status individually were not significantly associated with depression in the general U.S. adult population implies that both factors (i.e., being a non-citizen and having hypertension) are more likely to play significant roles in predicting depression when they interact but not in isolation.While the underlying factors are not clear, previous studies have reported that non-citizens face many structural barriers such as poverty, lack of health insurance coverage, under treatment, limited access to quality care, and non-adherence to prescription medications compared to U.S.-born citizens (Chen and Vargas-Bustamante, 2011;Guadamuz et al., 2020).A study which utilized the National Health and Nutrition Examination Survey (NHANES) to assess cardiovascular disease risk and treatment by citizenship status indicated that although non-U.S.citizens reported a lower prevalence of hypertension and other chronic conditions, they had considerably lower treatment rates than U.S.-born citizens (Guadamuz et al., 2020).Another study that used the Medical Expenditure Panel Survey and NHIS reported immigrants were less likely to take prescription drugs compared to U.S.-born citizens (Chen and Vargas-Bustamante, 2011).Similarly, the Korea National Health Insurance Claims Database was utilized to assess risk factors of non-adherence to antihypertensive drugs (Cho et al., 2020).The authors reported almost twice the level of nonadherence in immigrants with hypertension compared to native-born Koreans.Moreover, the risk of non-adherence was significantly higher among immigrants without usual source of care and younger adults (Cho et al., 2020).
Health status, sociodemographic and lifestyle factors also play a significant role in depression and other chronic disease outcomes  (Akhtar-Danesh and Landeen, 2007;Gabriel et al., 2021;Mann et al., 2022;Villarroel and Terlizzi, 2020).Accordingly, we found that current and former smokers, current and former drinkers, young adults, sexual minorities (i.e., non-heterosexuals), females, non-Hispanic Whites, those who were never married, unemployed, without health insurance, had low income, had more chronic conditions and had poor health were more likely to be depressed.Villarroel et al. analyzed symptoms of depression among U.S. adults, and, similar to our findings, they found young adults aged 18-29 years and women experienced the highest rate of mild, moderate and severe symptoms of depression (Villarroel and Terlizzi, 2020).Living in the U.S. for the longest time (i.e., >15 years) was also associated with higher odds of depression but this was not statistically significant in our study.For certain immigrant groups, length of stay may be an index of higher levels of acculturation which warrants further attention (Divney et al., 2019).Previous studies have suggested that acculturation plays an important role in depression, hypertension and other chronic conditions among migrants (Bernstein et al., 2011;Divney et al., 2019).
On the other hand, our findings indicate a decreased likelihood of depression among racial minorities and adults with low levels of education.Interestingly, insufficient leisure-time physical activity was also associated with lower depression, contradicting existing evidence (Fernandez-Montero et al., 2020).We recommend further studies that focus on several dimensions of leisure-time physical activity including the type, duration and intensity of these activities.

Clinical and research implications
The findings of our study imply that the healthy immigrant paradox, in the context of hypertension-depression prevention and control, may not apply to non-citizens with hypertension.While hypertension marginally increased the odds of depression among the general U.S. population, being a non-U.S.citizen with hypertension increased the risk of depression by 46 %.Thus, clinicians may need to pay more attention to these comorbid conditions, especially when presented among non-citizens.Community-based screenings, currently lacking among immigrant communities, should be encouraged since immigrants and other marginalized populations are less likely to access or use these services despite facing multiple stressors.There is also the need for researchers and clinicians to develop interventions that are culturally appropriate and sensitive to unique cultural practices and identities of specific immigrant sub-populations in the U.S.This can help improve cultural norms for health-seeking behavior and use of medication/ treatment among immigrants.Moreover, collaborative efforts are required to address barriers such as limited access to care and nonadherence to medications that are well-known to predispose immigrants to poor health outcomes as well as proper management of these chronic conditions (Cho et al., 2020;Hacker et al., 2015).We further recommend longitudinal studies and evidence-based interventions that take into consideration distinct sociocultural factors of immigrant communities in addition to targeting lifestyle, behavioral risk factors and access to healthcare.Henceforth, some policy changes may be necessary to effectively integrate immigrants into the health care system in light of its rapidly increasing and aging population in the U.S. (Bustamante et al., 2021).

Study limitations and strengths
Our findings should be interpreted with caution due to some limitations.NHIS data is cross-sectional, and therefore we cannot make causal inferences about our findings.Also, the variables we analyzed were self-reported, which may be susceptible to the respondent and social desirability biases and may underestimate our findings.Using complete case analysis to address missing or incomplete data can lead to bias and loss of precision in terms of estimations (Kang, 2013;Mukaka et al., 2016).However, the large sample size utilized in this study was very helpful in approximating accurate estimates and addressing biases in our results.Furthermore, we combined depression frequency and taking medication for depression to determine depression status, which may not be an objective, clinical, or valid measure of depressive symptoms.Nonetheless, this study still presents very important findings about depression among hypertension patients based on U.S. citizenship status since NHIS data is a validated nationally representative survey.

Conclusion
Non-citizens with hypertension had higher odds of depression compared to U.S. citizens without hypertension, underscoring the critical relationship between citizenship and hypertension in predicting depression.Thus, both citizenship and hypertension are more likely to The predicted marginals for proportion depressed are from a logistic regression model that included the following covariates: sex, sexual orientation, age, race/ethnicity, BMI, education, self-reported health status, income, years lived in the US, number of chronic conditions, leisure-time physical activity, smoking status, alcohol drinking status, health insurance, employment status, and marital status.Error bars indicate 95% confidence intervals.play significant roles in predicting hypertension when they interact but not in isolation.Community-based screenings and other interventions to address hypertension-depression among non-U.S.citizens should consider the unique cultural practices and healthcare barriers encountered by specific immigrant communities.

Fig. 1 .
Fig. 1.U.S. Adults with Depression by Citizenship Status and Hypertension, NHIS, 2015-2018.Note:The predicted marginals for proportion depressed are from a logistic regression model that included the following covariates: sex, sexual orientation, age, race/ethnicity, BMI, education, self-reported health status, income, years lived in the US, number of chronic conditions, leisure-time physical activity, smoking status, alcohol drinking status, health insurance, employment status, and marital status.Error bars indicate 95% confidence intervals.