Introduction

Medical professionals, health care providers, health promotion service providers, and researchers frequently consider the role chronic stress, and/or stressful life events can play in the lives and health of their patients or clients. Perceptions of stress are known to create feelings of anxiety and depression which can then influence behavioral patterns that have health consequences, namely clinical depression, cardiovascular disease, human immunodeficiency virus (HIV), and cancer [1]. However, not everyone exposed to stress experiences negative health consequences. Some individuals are more vulnerable to stress or stressful events than others and respond differently.

Multiple, as yet to be determined factors, such as conditioning, and consolidation of stressful memories, are suspected to influence persistent intrusive memories and chronic stress. These factors are the subject of continuing study [2]. Researchers have found that stress is linked, for example, to mental health status [3], smoking relapse [4], smoking attitudes [5], fat intake, exercise, alcohol consumption, and smoking behaviors [6], and being pregnant [7], and health disparities between racial/ethnic and socioeconomic subgroups in the population [8].

Those experiencing economic hardship are especially at risk for compromised mental and physical health and psychological stress [9, 10]. It would seem that learning to better cope with this stress would reduce psychosocial stress and consequently enhance one’s ability to cope with life’s stressors. Indeed, based on a randomized controlled trail, skills for coping and stress can be taught within a relatively short time. In the intervention trial cited here, a 14 h psycho-education intervention was effective [11].

Some researchers have concluded that there is a lack of comprehensive studies of the relationship between stress and health disparities for Latino/Hispanics in the United States, particularly those in poor and minority communities [8]. This population faces additional social or psychological stress because of the pressure to acculturate to the dominant culture. Acculturation researchers, for example, have found that English fluency pressures play a prominent role in the lives of immigrants or ethnic minorities and that needing to communicate in a new or secondary language is considered a risk factor for poor mental health—especially depression [12].

Based on this stated need, we report the findings of our study of Hispanic women aged eighteen or older living in the Border Region of South Texas who participated in a health promotion program called “Ready to Respond” (N = 284). Our goal was to determine variables associated with perceived stress.

Background

In clinical or research studies, patients, clients, or study participants are often administered a brief survey, such as Cohen and Williamson’s Perceived Stress Scale (PSS) [13], to determine their level of stress. The PSS is used to assess respondents’ perceptions of how unpredictable, uncontrollable, or how overloaded they feel [14], that is, how the respondents appraise their life as stressful and the degree to which they perceive their ability to manage that stress. The scale comes in three versions: 14, 10 items, and a short four- item scale. Based on an analysis of research articles using and or evaluating the various scales, the PSS-10 item scale was found to be superior to either the 14 or 4 item scales [15].

The level of perceived stress in the general population is also of interest to researchers [16]. Using data from three separate probability sample surveys (1983 Harris Poll Survey, and 2006 and 2009 internet-based, e-Nation Surveys) these researchers found consistent results over an almost 30-year time span: Perceived stress was higher among women compared to men, minorities tended to report more stress than whites (this relationship was not significant when education, income and employment were controlled for), stress decreased as age, education, and income increased, and perceived stress increased with unemployment (an exception was in 2009 where it did not increase). The mean perceived stress score for women in 1983 was 13.68 (SD 6.57), in 2006 it was 16.10 (SD 7.73), and in 2009 it was 16.14 (SD 7.56). In 2009, the most current survey, the mean score for Hispanics was 17.00 (SD 7.45), for Whites, 15.70 (SD 7.51), and for Blacks, 15.68 (SD 7.51).

To an extreme, one of the most vulnerable populations, those living with mental illness are generally considered to be at a greater risk for experiencing stress (perceived and objective) [3]. Based on the 14-item PSS, these researchers conducted a study of the validity of the PSS scale with institutionalized mental health patients who were close to discharge from the hospital after a brief admission. Using data from the MacArthur Study of Mental Disorder and Violence reported on in 2001 [17], they found the mean PSS score for all patients (N = 1136) was 37.92 (SD 7.23); for only the women patients, the mean score was 39.15 (SD 6.67.), and for the men, the mean stress score was 37.05 (SD 7.48). Scores for this population are significantly higher than the national average.

High Perceived Stress Scale scores among Hispanic populations are not necessarily unusual. In a study of pregnant Hispanic women in Massachusetts [7], the mean PSS score at early pregnancy (N = 979) was 26.2 (SD 7.1). The pregnant Hispanic women who indicated that they preferred to speak and use Spanish had 30 % lower odds of high perceived stress compared with those who preferred English. These researchers also found that stress levels significantly decreased over the course of pregnancy. Pre-pregnancy alcohol use and smoking were positively associated with high early pregnancy stress. Higher annual household income was inversely associated with high late pregnancy stress. Increasing age was positively associated with high early pregnancy stress.

Other researchers [5] have found that Latino light smokers, typically female, of Dominican or other Latino ethnicity, who were less nicotine dependent and who reported fewer “pros” of smoking had lower levels of perceived stress than heavier smokers (N = 131). These researchers used Cohen’s 4 item scale; however, no data was reported.

Methods

Participants

The study reported here is based on women 18 and over that live in a relatively poor community in South Texas where 27 % of all people in that community are considered in poverty, and where >95 % (95.7 %) are designated as Hispanic [18]. These women (N = 284) have participated in a health promotion program funded by the Substance Abuse and Mental Health Services Administration (SAMHSA), “Ready to Respond Initiative” (2010–2015). This grant was competitively awarded to applicants in areas highly impacted by substance use and HIV risk [19]. This program utilized an intervention called Coping With Work and Family Stress (CWFS) [20–22], an evidence-based intervention listed on SAMHSA’s National Registry of Evidence-based Programs and Practices [23]. Evidence-based programs have been researched and have demonstrated proven results. The program implemented a 12-session version of the intervention. Ninety minute sessions were provided to participants in one-on-one and in small-group formats, for a total of 18 h of the health promotion intervention. Sessions focused on helping participants understand stress, learning stress management techniques, and developing a personal plan to cope with stress [20]. Sessions were provided in either English or Spanish depending on the preferences of participants. CWFS has been shown to increase the use of active behavioral coping skills, the use of social support coping, reduction in avoidance coping (e.g., not doing anything), reduction in social withdrawal, and reduction in depression, anxiety, and somatic complaints. Two evaluation studies, with random assignment and controls, found that the intervention was effective in reducing substance abuse and psychological symptoms [22].

Data

Measures

Prior to program participation, all participants were asked to sign a “Consent to Program Participation and Services Form.” In this form, participants agreed to allow the program to share information collected with SAMHSA (the funding agency) for the purpose of doing a study to help prevent the occurrence of substance abuse and HIV infection. Participants were informed that only their answers to data collection instruments would be used and that no identifying information would be released.

Perceived Stress

All participants were administered the 10-item Perceived Stress Scale (PSS-10) prior to program services and post program services, approximately 3 months later. The PSS-10 assesses the degree to which life situations are perceived as stressful and measures perceived ability to manage stress [13, 14]. Scores range from 0 (never) to 4 (very often) for each item. Scores are then summed, with higher scores indicating greater psychological stress. Summed scores can range from 0 to 40.

National Outcome Measures (NOMS)

NOMS measures were created and mandated for collection by the US Department of Health and Human Services, Substance Abuse and Mental Health Services Administration (SAMHSA), Center for Substance Abuse Prevention (CSAP), for all CSAP Substance Abuse Prevention and Treatment Block Grant and Discretionary Grants [19]. From these measures, we selected those which most closely resembled the measures of researchers interested in the general population [16] and those that would be indicators of their Hispanic cultural orientation.

Demographic Variables

Educational level is measured by number of years of formal education. Age is calculated based on birth date. To measure employment, participants report if they are employed full-time, part-time, or unemployed (looking for work; disabled, doing volunteer work, retired, full-time student, full-time homemaker, other reason). Marital status is indicated as either single, never married, informally married or living with a permanent partner, legally married, separated, divorced or broken up from an informal marriage, and widowed. Income is measured at the household level and includes the income they and their family members made before taxes and ranges from $0–$10,000; $10,001–$20,000; $20,001–$30,000; $30,001–$40,000; $40,000–$50,000; $50,001; 50,001–$60,00; and more than $60,000.

Other Variables

Participants are asked if they have even been in juvenile/adult detention, jail, or prison for more than 3 days. Participants’ living arrangements are measured as living in their own home or apartment, in a relative’s home, in a group home, in a foster home, homeless or in a shelter, or other. Hispanic cultural orientation was measured two different ways. The first is how long have they lived in the US, categorized as less than a year, 1–2 years, 3–4 years, 5 or more years, or all their life. The second measure is language preference. Participants are asked to indicate their primary spoken language.

Analysis

Data

Of the original 284 cases that have both intake and exit PSS scores, nine were removed because participants indicated that they were “somewhat untruthful” or “very untruthful” (NOMS asks respondents this question). Another 16 cases were deleted because of significant missing data. The final number used for analysis was 259. The 25 cases eliminated represent less than 10 % of the cases (.088 %). SPSS was used as the analysis tool.

Results

Perceived Stress

At intake, the participants’ stress scores were very high. The mean score was 21.79 (5.75). The highest pre-PSS was 38 and the lowest was 1. At exit, approximately 3 months later, after the health promotion intervention, perceived stress scores decreased dramatically and significantly to a mean of 14.84 (S.D. 5.09). The highest score was 32 and the lowest was 0.

In Fig. 1 below, one can see that there was a “flip-flop” from pre to post. Pre-services scores by percent (e.g., the percent who had a score of 5, 6, etc.) were concentrated at the high end, above 20, and post-services score percents, were concentrated at the low end, below 20.

Fig. 1
figure 1

Pre and post PSS-10 scores by percent

Full size image

NOMS

Demographic Variables

Twenty-two percent of the participants had completed 12 years of school. Twenty percent had some post-secondary education. Almost 18 % had completed six or less years of education. The mean age of participants was 39.44 years. The youngest participant was 19 and the oldest was 79 years of age. Almost 26 % (25.9 %) were employed full-time (at least 35 h per week). Almost 12 % were employed part-time (11.6 %). The majority, 62.5 %, were unemployed. Of those who were unemployed, 23.9 % were homemakers. Almost a third (30.5 %) of the participants had never been married. Another third (32 %) were married. Almost 20 % (18.5 %) were living with a permanent partner, and almost 10 % (8.1 %) were widowed.

Other Variables

A quarter (25.1 %) of the participants had been in juvenile/adult detention, jail, or prison for more than 3 days. Almost three-quarters, 74.1 %, lived alone. Thirteen percent (13.1 %) lived with their mother. The other 12.7 % lived with relatives. Just over 40, 43.6 %, had combined family incomes of less than $10,000 a year. A quarter (25.1 %) had an income of between $10,001 and $20,000. Almost 13 % (12.8 %) had an income of between $20,001 and $30,000. Eight percent (8.1 %) had a combined income of $30,001 to $40,000. The remaining 10 % had incomes higher than $40,000. Almost 60 % (59.5 %) have lived in the United States all their lives. Almost 40 % (36.7 %) have lived in the US for five or more years. The majority, 62.2 %, consider Spanish as their primary spoken language, while 36.7 % consider English as their primary spoken language.

There are a number of correlated (and significant) variables. Length of time living in the US and educational level are both inversely correlated with age and language preference (Spanish). Employment status is inversely correlated with educational level. Jail time is inversely correlated with age and language preference, and positively correlated with time living in the US. Marital status is positively correlated with age and language preference, and it is inversely correlated with time living in the US, educational level, and employment. Home type is positively correlated with jail time. Income is positively correlated with educational level, employment, and inversely correlated with jail time and home type.

PSS at intake is not correlated with any of the variables (except the PSS at exit). The PSS at exit is inversely correlated with language preference (Spanish), educational level, and income, and it is positively correlated with employment and jail time.

Regression Model

Given that none of the variables were correlated with the pre-PSS, the regression model was run with the Post PSS scores as the dependent variable. The categorical variables were dummy coded. The regression model was significant at the p = .000 level (F = 4.976). All variables are in the expected direction (see Table 2).

Two of the predictor variables were significant at the .01 level or less. The first is language preference. Those who prefer Spanish as their spoken language are more likely to have lower PSS scores. The second is income level. Those with higher incomes have lower PSS scores. A third variable, employment, approaching significance at the .054 level, is of interest. Those who are employed tend to have higher post stress scores. The second variable approaching significance is education. Those with a higher education level tend to have lower post stress scores.

Discussion

Researchers examining Perceived Stress, nationally, propose that women and persons of lower SES are most likely to experience the highest levels of stress [16]. Our findings confirm theirs. At intake, mean PSS scores were very high, 21.79 (SD 5.75), almost a standard deviation higher than the national average score for women found by these researchers in 2009 (16.14, SD 7.56). This mean score is also higher than 2009 mean national score of 17 (SD 7.45) for Hispanics (both men and women). Likewise, lower family income among this study’s participants was significantly associated with higher PSS scores. It is noted that 43.6 % of the participants had combined family incomes of less than $10,000 and another quarter (25.1 %) had an income of between $10,001 and $20,000. Therefore, a total of 68.7 % of participants had an income of below $21,000. In the national samples [10], an income of lower than $25,000 was associated with higher PSS scores.

Income is often associated with employment and education. In this research, education level and employment were positively associated with income, and education level was significantly associated with lower PSS exit scores (see Table 1). Conversely, education level and employment were significantly inversely correlated with one another. Age and educational level also exhibited a significant inverse relationship. Thus, it appears that education may lead to higher income and higher income leads to employment, but associated with that employment is higher stress.

Table 1 Correlation matrix of variables
Full size table

We also found that those participants who preferred Spanish as their spoken language had lower stress scores (Table 2). In the community where the participants live, according to the US Census, 95.7 % self-designate as Hispanic and 91.2 % speak Spanish at home [24].

Table 2 Associations of variables with post PSS score as the dependent variable
Full size table

We find it striking that participants’ perceived stress scores declined so dramatically after 3 months of participating in a health promotion intervention. Again, scores declined from a mean of 21.79 to a mean of 14. 84. This program’s 18 h health promotion intervention appears to be effective in reducing perceived stress. Similarly, others have found success at increasing coping with stress with a 14 h intervention [11]. The program is not yet finished, and it remains to be seen if a reduction in perceived stress has a positive relationship with healthy attitudes and behaviors such as smoking, drug and alcohol abuse, and risky sexual behavior.

New Contributions to the Literature

As found in this study and another study of pregnant Hispanic women [7], initial high PSS scores declined over time; it was 3 months for our study participants. We suggest that intake scores are perhaps influenced by the new environment participants initially face; however, with familiarity over time to this environment coupled with services that met the clients’ needs, perceived stress declined. As found by other researchers interested in Hispanic immigrant health [25], a major social-ecological influence on health-care access and navigation of persons of Mexican descent living on the US/Mexico border is emotional fear or embarrassment. Unlike the participants in community public health initiatives, participants in the national surveys [12] were administered the surveys in the privacy of their homes, either over the phone or by internet with no further intrusion into their personal lives. We note that post services, the study participants’ PSS scores were lower than the national scores for women in the 2009 study (but within one standard deviation), and the national score for Hispanics (men and women combine) (but still within on standard deviation) [12].

Given that the intervention was successful in reducing Perceived Stress, ostensibly by teaching participants techniques and strategies to increase coping skills, perhaps those who are vulnerable to stress and consequent health risks, were preempted from engaging in risky behaviors, or were provided the skills to curtail existing ones. Since an extensive psychological battery was not provided as part of the program, there is no way of knowing who was at most risk. However, it does appear that the approximately three-month long intervention was able to change perceptions of stress.

Preferring the language that one speaks at home, in this case Spanish, is a significant predictor of reduced perceived stress. The number of people in the area who speak a language other than English at home is over 90 % [24]. This characteristic reflects a population that has minimally acculturated or is in the process of becoming acculturated to mainstream US culture, since language spoken at home is the most important indicator of acculturation [26]. The pressure to acculturate can be a major risk factor for many economically disadvantaged Hispanics, especially in regards to substance abuse and other at-risk behaviors [27]. Thus, the less that individuals have acculturated to mainstream culture appears to create a buffer against perceived stress.

Limitations

We suggest that the findings in the targeted poor and minority community are not likely representative of all poor and minority communities. There are limitations of this study. While it sheds light on the social, cultural, and economic aspects of Hispanics, it is noted that the community in which this study’s participants reside is not a minority community within a majority community, but rather where the minority is the majority. The dynamics of race or ethnicity are likely very different than in other places in the United States. Additionally, correlation/association is not the same causation. While low income was associated with higher perceived stress, for example, a significant percent of this study’s participants had incomes much lower than the participants in the national surveys. How low is “too low?” We propose that income needs are relative, that there may be other protective resources available to Hispanics living in a predominately Hispanic environment. Learning how to access and utilize these may be instrumental in enhancing coping with low income or acculturation pressure.

It appears that a relatively brief intervention, and in this case an evidence-based intervention, can make a significant difference in perceived stress levels of participants, that attitudes and behaviors can change for the better. Whether these changes hold for the long run, that is, what participants have learned is used in their futures, is something that deserves investigation.