Introduction

The co-morbidity of diabetes and depression has been reported as high as 33% in Latinos served in primary care [1]. This population faces considerable disparities in accessing and receiving high quality care for these conditions [2]. Compared to non-Latino Whites, Latinos are 1.5 times more likely to have diabetes and twice as likely to develop diabetes related complications than other populations [3]. Latinos are less likely to seek care for such complications and when they do receive care it tends to be poor in quality [2, 4–6]. Diabetes among Latinos leads to higher functional impairment and mortality rates compared to non-Latino Whites [3, 7]. Despite these inequities in care, little is known about how Latinos with diabetes recognize their need for depression care, initiate professional treatments, and adhere to depression treatments. Addressing this under studied area can identify factors and processes that contribute to disparities in care and inform the development of culturally appropriate interventions aimed at increasing service use, adherence to care, and treatment effectiveness.

Contributing to the inequities in service use and provision of effective treatments is that Latinos receiving treatment for diabetes are not adequately screened for depression and receive low rates of treatment when co-morbidity is diagnosed [8]. Seeking treatment for diabetes and depression is complicated by socioeconomic factors (e.g., lack of health insurance, inability to pay for care) creating decreased access to care and resources to advocate for health needs [9–11]. Support systems including family or social service providers can bridge divides in available information on appropriate care and related services, such as nutrition, education, and medication management. Language barriers are another form of informational divide for many immigrant Latinos with limited English proficiency (LEP). Latinos with LEP face numerous obstacles in accessing and understanding health information, navigating the service system, and communicating their needs to providers without interpreter assistance [12, 13].

As stressed by the Surgeon General’s [4] report, culture counts in explaining patterns of service utilization and treatment for ethnic minorities. Abundant research in medical anthropology and health psychology indicates that cultural norms, cultural perceptions of illness and treatment as well as stigma influence help-seeking decisions, treatment initiation, retention in care and management of health conditions [14–17]. For example, the cultural value of familism, which places the family at the center of Latino individuals’ identity and everyday life, shapes Latinos’ help-seeking decision by relying on family for instrumental and emotional support during time of stress [19].

Untreated symptoms resulting from co-morbid depression and diabetes can impair functional status inhibiting work, self-care, and social relationships, especially for Latinos who report high rates of co-morbid conditions [18]. The burden associated with symptoms from both illnesses is exacerbated by delayed initiation of care and adherence to care. Diabetes management is a complicated process for patients who must balance diet, exercise, medication treatment, and glucose monitoring. Combining this intensive regimen with depression makes it difficult for patients to maintain the motivation to achieve optimal outcomes [20]. Many individuals do not seek care until they experience acute symptoms that impair functioning [9, 19]. These delays in care may be more pronounced among low-income Latino immigrants who lack a regular source of medical care and face numerous barriers in accessing and receiving adequate treatments for depression and diabetes [6, 13].

In the present study, we used a combination of focus groups and in-depth qualitative interviews to examine the help-seeking pathways of low-income Latinos with diabetes in managing their depression. There has been little examination of these pathways reported in the literature [22]. We explored the social, cultural, and interpersonal factors that influence help-seeking and barriers in managing depression and receiving healthcare in primary care.

Methods

The University of Southern California Health Sciences Institutional Review Board approved all study procedures. Study methods have been reported elsewhere [22] and are briefly summarized. The study sample was drawn from a randomized controlled trial (RCT), the Multi-faceted Depression and Diabetes Program for Hispanics (MDDP) that examined the effectiveness of a collaborative depression care model adapted for low-income, Latino adults with diabetes at two large public primary care clinics in Los Angeles, California. Both clinics are deemed safety net clinics where the patient populations are predominately Medicare and Medicaid recipients. Though income data was not collected for the purposes of this study, income data from the MDDP parent study cites 78% of the MDDP sample reported having no money left over at the end of the month to cover expenses, 46.5% reported feeling very worried about current financial situation, and 78% of the sample was unemployed. These factors indicate an overall low-income population [24]. MDDP aimed at reducing depressive symptoms and improving patient adherence to diabetes self-care regimens, glycemic control, functional status, and quality of life for Latino populations with limited resources and access to mental healthcare. Participation in the RCT was based on having a diagnosis of diabetes and screening positive for major depression. Diabetes screening occurred through a review of patient medical charts. Depression screening was done through administration of the Patient Health Questionnaire 9-item index (PHQ-9) [26] by a trained screener for the MDDP program [25]. Participants were ineligible if they presented with suicidal ideation, had a primary language other than English or Spanish, indicated alcohol abuse, reported recent use of lithium or antipsychotic medications, and presented with cognitive impairment.

Recruitment Procedures

To explore the unique help-seeking experiences of Spanish-speaking Latinos, participants were recruited for this qualitative study if they were enrolled in the parent grant by August 2006, were Spanish-speaking Latinos assigned to either the intervention (INT) or usual care (UC) groups and had completed baseline interviews. Intervention group patients were required to complete at least one visit with their Depression Diabetes Care Specialist in order to be eligible for the present study. This eligibility criterion allowed us to include individuals who received depression care by the time of our interviews. In these initial visits, intervention patients received patient education materials about depression, completed a psychosocial evaluation with their therapist, and initiated depression treatments (e.g., antidepressant medications, problem-solving therapy) [26]. UC patients received referrals to community mental health resources and were free to receive any type of depression care already available through their primary care physician at the clinic. Fifty-seven patients were invited to participate, 28 (49%) were from the usual care group; 32 agreed and 25 refused participation. Of those 32 individuals who agreed to participate, 19 (59%) attended the focus groups. Eleven were in the INT and 8 in the UC group. Within the final sample, thirteen participants had received some type of depression treatment by the time they were interviewed. Nine received antidepressant and counseling; two received antidepressants only; and two received counseling only. Comparisons of participants and non-participants and between INT and UC patients found no statistically significant differences or trends in age, gender, years of education, or marital status. Depression status and baseline depression levels were measured with the Patient Health Questionnaire 9-item index (PHQ-9) [25]. Of the whole sample, INT and UC, four focus groups were conducted; 2 from INT and 2 from UC groups.

Focus groups participants were invited to participate in follow-up semi-structured in-depth qualitative interviews. Five focus group members declined to participate. We screened out 4 individuals from participating in the individual interviews due to clinical decomposition during the focus group discussions. Clinical decomposition in this case refers to participants who once exposed to questions about their experiences with depression became uncontrollably emotional and unable to participate in the group process. These individuals were referred for individual follow-up with a trained therapist affiliated with the study. A total of 10 individuals participated in the individual interviews, 5 from UC and 5 from INT groups. Participants received gift cards valued at $15 for their participation in the focus groups and in-depth individual interviews.

Data Collection: Focus Groups and Interview Methods

Focus groups were held at the participants’ primary care clinics and lasted 90 min with 3–6 participants in each group. Trained bilingual researchers facilitated all groups. An interview guide informed by Kleinman’s [16] framework of explanatory models and Leventhal et al. [27] common sense model of illness was used to explore the following areas: (1) explanatory models of depression (2) perceived relations between diabetes and depression (3) help-seeking behaviors for depression, and (4) depression treatment experiences. Questions included those related to how the participants manage their diabetes and depression, how they sought help for depression, attitudes regarding their treatment for the illnesses, social norms/stigma around their experience with depression, and desired outcomes for their own physical and mental health. Questions were asked about treatment experiences overall and not only specific to the MDDP. Researchers obtained written informed consent from each participant. In this study, we only report findings from our examination of help-seeking behaviors for depression. Findings from the other themes are presented elsewhere [23].

Seven individual in-depth interviews were completed in-person and three by telephone, each lasting 60 min. Difference between the phone and in-person interviews were deemed minimal given use of standard interview guide and training of interviews on interview techniques using both mechanisms. Construction of the interview guide was an iterative process based on themes that emerged from preliminary analysis of the focus groups. Total sample used for data collection, analysis, and presented in this paper is pooled from both the focus groups and individual in-depth interviews.

Data Analysis

Digital recordings and professional transcriptions of all focus groups and in-depth interviews were completed. Trained research assistants conducted a cross comparison of the recordings and transcripts in order to correct errors in transcription. Using an analytical approach based on grounded theory [28], data analysis incorporated coding, consensus, co-occurrence, and comparison methods [29]. Analysis of focus group interviews provided the basis for an initial codebook with new codes added from the review of individual interviews. Codes were developed with first and second authors working independently to review transcripts based on a priori and emerging themes. We wrote analytical memos describing definitions of codes and decisions during analytic process, highlighting illustrative text segments for each code. The authors met weekly to debrief, discuss interpretations, resolve disagreements through consensus, refine and identify new codes. The finalized codebook emerged from this iterative process. We conducted line-by-line coding with Atlas. ti [30] software for all transcripts and drafted reports, queries, and memos to compare findings from both focus group and individual interviews. This enabled us to further the process of axial coding to identify associations between a priori and emergent themes. By constantly comparing these categories with one another, the different categories were condensed into broad themes that were linked into a heuristic model. We also explicitly examined whether responses from focus groups and individual interviews differed in regards to the categories and themes that formed the basis for our heuristic model. No differences were found between these two sets of data, thus results are presented from this pooled data.

Results

Sample

Sample characteristics are summarized in Table 1. The 19 participants had a mean age of 57 years, were predominately female, and of Mexican decent. Participants lived an average of 25 years in the U.S. and none were U.S. born. Depression levels (PHQ9 score) ranged from moderate to severe (mean = 15.3, SD = 2.45) at the time participants were screened into the RCT. All individuals were diagnosed with Type 2 diabetes and had an average baseline Hba1c level was 8.36% with a large portion reporting poor glycemic control, based on the American Diabetes Association guidelines (HbA1c above 7.0%) [31].

Table 1 Sample characteristics (N = 19)
Full size table

Three distinct themes organized within the help-seeking framework emerged from participants’ narratives about how they sought help for depression: (1) recognition of need for care (2) treatment initiation, and (3) adherence to care (See Fig. 1). Themes emerged across the participants within the UC and INT group and were not unique to treatment provided by the RCT in this case [22] as both group had been exposed to depression treatment by the time they were recruited into this study. We include a section discussing barriers to care across these help-seeking steps. All patients’ quotes have been translated into English using forward back translation by trained bilingual research staff. All text was reviewed by authors who are both English/Spanish bilingual to assure that consensus was achieved in the translation process.

Fig. 1
figure 1

Help-seeking pathway model

Full size image

Recognition of Need for Care

Recognition of need for formal care was described in relation to somatization of depressive symptoms (e.g., fatigue, chest pains), functional impairment and/or mood changes. Most striking were reports that associated symptoms with decreased functioning. These included inability to work or care for family members, and disruptions in interpersonal relationships. A participant discussed how somatic-like symptoms led her to seek care from her primary care doctor.

Doctor, I feel like I have something here [touching her chest], like a car, like it accelerates me… I feel very bad. (Focus group)

In many cases, participants discussed how they delayed talking with anyone about their symptoms and relied on self-management techniques, such as avoidance, denial, isolation, and prayer. As their symptoms deteriorated and coping techniques were no longer successful, participants described how they realized they were in need of formal help. For some, acquaintances or family members facilitated recognition of a potential mental health problem by noticing and communicating concerns about participants’ behaviors. These comments prompted participants to realize there was something wrong and that they needed to seek help, as illustrated by this female participant

I would get up with that depression, much sadness and I would cry….and my daughter said to me one day “Listen mom… how come all you do is cry…?” She said, “You never cry…” A girl of eight years told me that. I started thinking…I have to look for help because this is making them feel bad and me also. (Focus group)

For others, the recognition of need was prompted by first seeking help from a priest or religious leader. As one focus group participant commented, her priest was the first individual to encourage her to seek psychological help. Lastly, participants discussed how interactions with clinic staff and study recruiters from the parent grant who screened for depression were notable experiences in recognizing their need for care. For many this was the first time someone asked them about their depression and provided them an opportunity to talk openly about their feelings and symptoms.

Treatment Initiation

Participants moved from recognizing depressive symptoms to actively initiating a form of treatment. Family members motivated participants to take the initial steps to access care through encouraging visits to their physician, prompting participants to become more active and advocate for their care needs. Without the external motivation from family, participants expressed resistance to access services For instance, one focus group respondent related how she was urged by her husband to talk to a female doctor in hopes this would help her feel comfortable discussing health concerns, which the participant struggled to do until that point.

For many, the initial contact with their physician facilitated the acceptance and understanding of their depression. In this first visit, participants related how their doctors were instrumental in helping them come to terms with their illness and decision to receive treatment, as exemplified by the comments of this female participant:

Like the doctor tells me “You have to accept your diabetes. You have to accept your high blood pressure. You have to accept….bad moods….like you accept your problems, you have to accept your illness”. And I now, that is what I am trying to do, accept. (Focus group)

Another important aspect valued by participants in these initial visits was the empathy they received from their doctors, highlighting the relevance that doctor-patient relationships had at this stage of the help-seeking pathway.

Adherence

Thirteen participants received depression treatments (medications and/or psychotherapy) by the time they were interviewed. Three processes for engaging participants into care and adhering to treatments were identified. First, respondents recounted in both focus groups and individual interviews how positive interactions with their physicians and therapists contributed to their adherence. Participants valued providers who were nonjudgmental, warm, and instilled confidence that they could overcome their depression. A common sentiment was feeling that providers genuinely value participants as human beings. For many, the development of a trusting relationship with their provider was essential in adhering to treatment.

Second, participants described how noticing improvements in symptoms heightened their awareness that the treatment was working. Relief from symptoms, such as lack of energy, decreased appetite, anxiety, and sleep disturbances, provided participants tangible evidence that treatment was working. For some, this realization motivated them to continue taking their medications or return to their psychotherapy sessions. Others interpreted this relief as not needing further treatment, which resulted in self-medicating practices or prematurely discontinuing treatment. Self-medicating practices by patients were discussed as modifying dosage levels of anti-depressants or skipping days based on how they felt or if they were concerned about effects of medication on their body.

At times I take a half of the pill for depression…it’s what helps me, it’s what calms me…. I don’t take it every day, only when I get to the point of feeling a lot, a desire to cry, with anxiety… (Individual interview)

Third, participants described how improvements in functioning (e.g., return to work) reinforced their adherence to treatment. This was most notable among participants who expressed relief from impaired functioning as an impetus for seeking care.

Barriers to Care

Several barriers to care were identified across the three help-seeking stages. One of the most prominent issues discussed were language barriers. These language difficulties prevented participants from clearly expressing their symptomatology, conveying their need for care and understanding providers’ instructions for treatment. Participants recognized that these gaps in communication impacted the quality of the clinical encounter. Even with the assistance of interpreters, participants felt that their communication with providers was fragmented, as expressed by the comments of this participant.

Look, in reality with the doctor here, I cannot communicate well, because he does not speak any Spanish. The doctor is informing me through an interpreter and so the interpreter does not tell us everything, because one very clearly sees… she is listening and in the whole time tells us two, three words. So realistically, you understand, that it is not everything that the doctor is saying. (Individual interview)

Another common barrier was the stigma associated with taking medications. Many participants resisted and feared taking antidepressants because they equated these medications to treatments used only for individuals they termed as locos (crazy).

Well, I heard that it makes you sick, like that is the medicine for when one is loco [crazy] and all that…Well, also because of that I didn’t want to take them, because I was scared they would make me ill in the head… (Individual interview)

Participants also viewed antidepressants as addictive and harmful to their bodies. These attitudes negatively impacted adherence.

Systemic barriers in accessing services were commonly noted. Participants expressed frustration with obtaining medical appointments, and long waiting times to see a doctor. Short visits with doctors were also a major concern and hampered patient’s ability to adequately describe symptoms and discuss their need for care. They reported leaving their visits with unresolved health and mental health concerns and questioning whether they could come back to their providers for additional care.

Well, since they treat me every six months… we hardly have talked, I only come and they look me over, and they say to me “where does it hurt, if it hurts”. They only prescribe me the medication and that is it.

Interviewer: You have never thought of talking to your doctor?

Respondent: No, because almost it never lends itself to do so, or moreover, since I know there are many patients, they try to get them out fast. (Individual interview)

When participants did speak with a doctor, some reported at times receiving limited information on available medication or mental health services from their doctors. Instead, patients reported that physicians provided advice on coping techniques (e.g., breathing exercises, behavioral activation).

Lastly, participants reported that managing multiple demands (e.g., caregiving burden, domestic violence, co-morbid health conditions) impacted their ability to initiate and adhere to depression care. Certain participants described not having the time to attend therapy appointments due to work or family responsibilities. Others seemed to prioritize health concerns over mental health problems thus delaying the receipt of mental health treatments.

Discussion

Help-seeking behaviors for depression revolve around three steps, recognition of need, treatment initiation, and adherence to care. Each step highlights key facilitators and barriers that impacted participants’ decisions in managing depression within primary care settings. As stipulated by Rogler and Cortes [22], help-seeking does not happen in a vacuum detached from the sociocultural environment of the individual. Help-seeking for depression is impacted by a complex and dynamic interaction of individual, social, and organizational factors.

Recognition of need for care and understanding the tangled relationship between somatic and emotional symptoms of co-morbid disorders was the most prolonged process in the pathway to care. Participants discussed their initial attention to somatic symptoms and lack of understanding of the cause of their impairment. Somatic complaints are common symptoms for individuals experiencing depression and chronic illnesses [19, 32]. Participants noted that symptoms negatively impacted their ability to work, complete daily tasks, and strained personal relationships. Symptoms of depression and diabetes individually produce a psychological reaction to the physical disease which becomes exacerbated by the co-morbidity [33]. The acute experiences of these symptoms, along with impaired functioning, prompted some participants to recognize the need to seek help from their doctor. Participants with this type of experience present a direct pathway in which help-seeking is prompted by attention to acute symptoms and reductions in functioning.

Others resisted seeking formal care and exhibited self-reliant attitudes. Participants reported coping with symptoms on their own by engaging in self-management techniques, such as avoidance, denial, isolation, and prayer. Self-reliant attitudes are linked to delays in seeking formal mental health services [34]. For these participants, such techniques extended their depressive symptoms and functional impairment, contributing to poor self-care and triggering acute needs [19] and subsequent help-seeking.

Consistent with previous studies [35–38], family and friends were important in the recognition of need and treatment initiation. These informal sources of support called attention to changes in participant’s behavior or emotions. However, others reported feeling unable to talk to anyone about their problems. The hesitancy to discuss their feelings and condition was fueled by uncertainty about the cause of their symptoms, negative attitudes towards mental illness, or not wanting to worry family about their needs, all which delayed treatment initiation. For participants who expressed inhibited help-seeking within their immediate network, alternate support (e.g., priest, doctors) was sought. These participants experienced a less direct path to treatment that required them to expand help-seeking efforts outside intimate social networks. The new contacts acted as informational bridges in helping identify their need for care and initiate treatment [21]. Reported findings reflect how for many low-income Latinos the cultural and social milieu play an important role on help-seeking decisions and illustrate how help-seeking begins within family unit and expands out when the situation deteriorates or care within this intimate network is exhausted or not provided [22, 35].

Central issues in adherence to treatment focused around interpersonal aspects of care, appraisal of symptom relief and improvements in functioning, and stigma. For our participants, having a trusting relationship with their provider characterized by respect and warmth seemed to enhance motivation to take antidepressant medications and/or attend therapy sessions. These findings reflect salient cultural norms, such as confianza and simpatía, that Latinos value in interpersonal relationships. Studies show that the quality of patient and provider relationships increase patient’s abilities to manage their chronic health condition and adhere to care [2, 18, 39]. Incorporating these cultural norms into depression treatments, particularly in the early stages of treatment, can enhance the cultural relevance of care and improve patient retention and adherence

Participants’ evaluations of how their treatment contributed to alleviation of symptoms and improvement in functioning were linked to adherence behaviors. This appraisal motivated them to adhere while for others it suggested that treatment should be modified (e.g., only take medication when depressed) or no longer needed. These findings suggest that individuals’ subjective experiences of treatment, particularly perceived treatment benefits, evolved through the course of treatment and played a role in shaping adherence. Clinicians’ attention to how patients appraise the impact of treatment can help identify maladaptive behaviors and perceptions to prevent non-adherence [19, 20].

As reported in previous studies [40], many of our participants viewed antidepressants as addictive and harmful and feared taking these medications because they did not want to be labeled as locos (crazy). Stigma negatively impacts treatment retention among older depressed patients [41] and adherence to antidepressant medications [42]. This issue also can perpetuate self-reliant attitudes and subsequent delays in care towards given fears about disclosing mental health needs [33]. Despite the widespread stigma surrounding mental illness in the U. S., little empirical attention has been given to the stigmatizing attitudes toward common mental disorders and treatments in Latino populations, particularly immigrants and those with limited English proficiency. Our study contributes to emerging findings that stigma may partly account for the underutilization of services in ethnic minority groups [17, 43, 44] and impact treatment adherence once in care. More research is needed to design linguistically and culturally appropriate interventions to reduce stigma and promote help-seeking for depression in Latino communities.

A constellation of barriers were mentioned throughout the pathway to care but appeared most pervasive in the initiation and adherence steps. Previous research reports that variability in treatment experiences exist when patients are unable to communicate their needs in their native language [9, 15]. Latino patients with limited English proficiency served in primary care may benefit from Spanish speaking providers or increase integration of trained interpreters who actively engage patients in interactions with physicians [45]. Such experiences can enhance patients’ understanding of their health conditions and treatments, assist in open discussions about treatment concerns and fears, and help develop better self-management skills to effectively cope with their chronic illnesses. Though this sample reported on average 25 years of residency in the United States, they were all mono-lingual Spanish speakers thus impacting their pathways and use of care. Reductions in language barriers are an essential component for improving the quality of health care and outcomes of populations with limited English proficiency. Language and culture in this context become intertwined with how patients communicate individual preferences for care and how providers communicate with patients in addressing their needs. Both are influenced by personal beliefs, experiences, and even biases about expectations and decisions for care [46]. Additional patient level barriers need to be addressed, including limited time with providers, difficulty in scheduling appointments, and long wait times in the clinics. The strained context in which patients seek care, particularly in busy safety net clinics, activates help-seeking attitudes that lead patients to prioritize what they believe to be their most acute needs.

Findings illustrate influences and barriers throughout the pathway to care and highlight potential points of intervention. Community outreach and education on the risk factors and symptoms of depression can increase awareness and promote treatment initiation. Clinicians should account for social and medical factors that influence treatment of depression and diabetes to facilitate care and maximize treatment outcomes [33]. Diabetes is a demanding disorder impacting lifestyle and functioning over time, which places individuals at risk for increased health and mental health complications when not appropriately treated, especially for Latinos [47]. Screening for depression in primary care and providing integrated health and mental health treatments can assist in increasing rates of treatment for both illnesses and functions as a preventive measure to address the negative effects of depression on this chronic disease [47]. Cultural and linguistic adaptations of collaborative care approaches to treat depression and diabetes within primary care can enhance the congruency between Latino patients’ expectations and treatment goals thus improving quality of care, rates of service use, and treatment outcomes in this underserved population.

Results of this study have limited generalizability to other low-income Latino populations with diabetes and depression. Participants were part of an RCT study based at two large, public primary care clinics in Los Angeles, California and were predominately of Mexican descent, female, and foreign-born. Therefore, help-seeking experiences of other Latino populations in the U.S. may not be accurately represented, especially among those not undergoing treatment for diabetes or depression. Further, the sample was comprised of a population already receiving some form of medical care through affiliation with primary health care clinics and/or the RCT. The sample does not fully represent the experiences of those who do not have access to regular sources of care. Future studies should also include participants from general population outside of systems of care, and individuals from Latino sub-groups, English speaking Latinos, and a larger portion of males to best represent the heterogeneity of the Latino population. Including a comparison group, such as non-Latino Whites, would also enrich the understanding of which factors are specifically relevant to Latinos compared to other populations. Further including English speaking individuals, Latinos and non-Latinos, is important given that English speaking individuals have access to resources and opportunities for seeking help and engaging in care that may not be available or hard to reach LEP populations. The combination of focus group and in-depth interviews enabled us to elicit rich contextual information on the help-seeking experiences of the sample which is rarely discussed in the literature. Results from this study can inform future research that allows for hypothesis generation and theory development of the psychosocial and cultural influences of Latinos with co-morbid health and mental health disorders. In all, this qualitative study highlights potential points of intervention for developing culturally-appropriate collaborative care approaches for low-income Latinos with diabetes and major depression.