nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
By the year 2020, cardiovascular diseases (CVDs), depression, and anxiety are projected to become the leading causes of disability and loss of productivity1. They are among the most prevalent diseases worldwide and frequently coexist2. In addition, anxiety and depression often take place simultaneously; patients with depression have 47-56% more probabilities of developing anxiety in their life, while 56% of the patients suffering from anxiety will develop depression3.
Furthermore, depression and anxiety have been recognized by several studies as risk factors for the development of CVD4-7. For decades, mortality associated with depression has reached over 40% in patients with CVD8,9. In fact, the presence of depressive and anxious symptomatology among individuals with CVD has been associated with poor prognosis and a higher risk of mortality independently of disease severity, depression level, or lifestyle1,10-12. However and despite the negative impact of the co-occurrence of depression and anxiety in CVD patients, diagnosing depression occurs in less 15% of the cases as pointed out by a study carried out in Poland3. Furthermore, a study conducted with 74 patients with acute myocardial infraction found that 69% and 50% of the patients had anxiety symptoms and anxiety disorder, respectively; however, they were missed by the health provider13.
Due to the high prevalence of these diseases and the overlapping between them, it is extremely important that health providers can recognize such conditions to improve the patients prognosis by giving them the correct treatment. The hospital anxiety and depression scale (HADS), a self-report instrument which does not require special training for the application and interpretation, is widely used to screen anxiety and depression in patients with different medical conditions because it avoids the overlapping of physical symptoms shared by such emotional and health problems. The scale is integrated by two subscales, each one with 7 items; one for the evaluation of anxiety HADS-A and other for depression HADS-D2,14. Multiple studies have demonstrated the good psychometric properties of the scale in different settings and with different population2. A review conducted to assess the validity of the HADS concluded that the scale is useful for the detection and severity evaluation of anxiety and depression in a variety of populations15. Studies which evaluated the psychometric properties of the scale in patients with CVD also endorse its reliability13,16.
In Mexico, there have been several studies assessing the psychometric properties of the HADS Spanish version in different samples, including patients with mental disorders (i.e., eating disorders)17, physical illnesses (such as cancer and HIV)18,19, and health issues such as obesity20, as well as in cancer caregivers21. All of them have proven that HADS is characterized by good psychometric properties for the identification of depression and anxiety in Mexican population.
Recently, Herrera, Galindo, and Lerma in 201822 evaluated the HADS in Mexican patients with CVDs, finding a one factorial structure that differs from the original version of the scale with two factors; one for depression and the other for anxiety14. The aim of this study is to evaluate the psychometric properties of the HADS Mexican Spanish version in a bigger and without cognitive deficit or psychiatric disorder sample of patients with CVD, specifically in individuals with ischemic or hypertensive cardiomyopathy.
Materials and methods
Study population
Recruitment was performed by a convenience sampling approach from adult outpatient services of 12 different hospitals of the second and third level of attention in cardiology throughout the Mexican Republic from March 9, 2016, to January 1, 2017.
Participants were approached in the waiting area and interviewed by trained staff in the participating hospitals if they met the following inclusion criteria: (1) 18-75 years old, (2) having an ischemic or hypertensive cardiomyopathy diagnosed by a cardiologist, and (3) able to read and write in Spanish. From the 1954 patients approached to the hospitals, only 953 met the inclusion criteria. Participants were excluded from the study if they had cognitive deficit, decline, or any major psychiatric conditions such as psychotic or bipolar disorders. Patients who agreed to participate were given the HADS scale, while they were in the waiting room. Information on cardiovascular risk factors was extracted from the patients medical record where available. Each participant signed an informed consent to enter into the study. The research project was approved by each hospitals ethics committee and Internal Review Board.
Study instruments
A detailed sociodemographic questionnaire was generated to collect information on age, gender, level of education, marital status, and employment. Clinical features related to the cardiomyopathy such as illness evolution, history of heart attack, previous surgery, current medications, and comorbid medical conditions were registered. The patients CVD was documented, assigning them to follow hypertensive heart disease or ischemic heart disease.
The Spanish version of the HADS was used to measure symptoms of depression and anxiety. This scale has 14 items with two subscales; one for depression (pair numbers) and another for anxiety (odd numbers). Each question is scored on a Likert scale from 0 to 3, whereby 0 is the lowest and 3 the highest severity level of anxiety and depressive symptoms. The final score is obtained by adding up the values of each subscale. Results can range from 0 to 21. A cutoff point of 11 was used to define a positive screening14.
Statistical analysis
The procedures for testing the psychometric properties of the HADS scale were done as follows: (1) the item discrimination index23 with the method of extreme groups24 was obtained for each item of the HADS; items with moderate indexes (less 0.30)23 were eliminated; (2) item correlation was tested and those with a correlation coefficient lower than 0.20 were eliminated; (3) construct validity was determined with an exploratory factor analysis with varimax rotation; items with communalities greater than 0.40 were retained and allocated to factors; and (4) the factor congruence coefficient25,26 of the HADS factors were determined using the results of the HADS obtained in a Mexican sample of cancer patients18, where a same bifactorial structure was found. Coefficient values > 0.60 were considered as congruent. All analyses were the SPSS-X version 20 for Windows, PC.
Results
Sample description
A total of 953 individuals with ischemic (n=562, 59%) or hypertensive (n=391, 41%) cardiomyopathy were included. The mean age was 58.3 years old (SD=11.1, range 18-75), with more than half of the participants being men (n=573, 60.1%), where 26.9% (n=257) had secondary school, followed by 24.4% (n=233) with elementary school, 23.8% (n=226) had a bachelors degree, 21.8% (n=207) completed high school, and the remaining 3.1% (n=30) had postgraduate studies. At the time of the study, 74.2% have a couple (married 663, 69.6%; and cohabiting couple 44, 4.6%) and 51.9% (n=495) had a remunerated employment.
Mean evolution of the cardiomyopathy was 6.5 years (SD=7.2, range 1-49 years) and more than 50% had at least a surgery related to their heart condition (52.9%, n=504). Hypertension and diabetes were the most frequent comorbid conditions, and patients were under several medications when the interview was performed (Table 1).
Table 1 Clinical characteristics of the sample
| Cardiomyopathy features (n, %) | |
|---|---|
| Illness evolution (years)* | 6.5; 7.2; 1-49 |
| History of heart attack Yes | 331 (34.7) |
| History of surgery Yes | 504 (52.9) |
| Years elapsed since the last surgery* | 3.3; 4.2; 1-36 |
| Main medical comorbidities (n, %) | |
| Hypertension yes | 719 (75.4) |
| Diabetes yes | 369 (38.7) |
| Renal failure Yes | 28 (2.9) |
| Dyslipidemia yes | 262 (27.5) |
| Respiratory diseases yes | 38 (4.0) |
| Chronic obstructive pulmonary disease yes | 21 (2.2) |
| Asthma yes | 16 (1.7) |
| Both yes | 1 (0.1) |
| Current treatment (n, %) | |
| Acetylsalicylic acid yes | 612 (64.2) |
| Beta blockers yes | 476 (49.9) |
| Clopidogrel yes | 274 (28.8) |
| Statins yes | 511 (53.6) |
| Hypoglycemic agents yes | 279 (29.3) |
| Insuline yes | 122 (12.8) |
| Anticoagulant yes | 70 (7.3) |
| Angiotensin-converting enzyme inhibitor yes | 456 (47.8) |
| Antithrombotic yes | 17 (1.8) |
| Calcium blockers yes | 162 (17.0) |
Item discrimination indices and item correlation coefficients
The item discriminant indices for the 14 items of the HADS range from good to excellent with values between 0.370 and 0.857. Furthermore, the item-total correlation coefficients of the items were adequate, ranging from 0.477 to 0.648. Therefore, all the items of the HADS were included in the subsequent factor analysis.
Factor analysis and internal consistency of the HADS
The results of the varimax rotation of the 14 items of the HADS accounted for 50.5% of the variance. The scree plot showed two clear factors that correspond to the original two designated domains of the HADS, named anxiety and depression (Fig. 1).
In addition, each of the 14 items loaded into the factor for which it was designed. None of the items loaded into more than one of the factors. The Cronbachs alpha for the combined items of the HADS 14 items was 0.88, with a high reliability also obtained for the anxiety and depression factors (Table 2).
Table 2 Exploratory factor analysis and internal consistency HADS 14 items
| Items | Anxiety factor | Depression factor | |
|---|---|---|---|
| A13 | I get sudden feelings of panic | 0.773 | 0.217 |
| A1 | I feel tense or wound up | 0.724 | 0.205 |
| A5 | Worrying thoughts go through my mind | 0.707 | 0.164 |
| A3 | I get a sort of frightened feeling like something awful is about to happen | 0.669 | 0.258 |
| A11 | I feel restless as if I have to be on the move | 0.652 | 0.186 |
| A9 | I get a sort of frightened feeling like butterflies in the stomach | 0.538 | 0.272 |
| A7 | I can sit at ease and feel relaxed | 0.519 | 0.275 |
| D14 | I can enjoy a good book or radio or TV program | 0.147 | 0.906 |
| D2 | I still enjoy the things I used to enjoy | 0.147 | 0.906 |
| D4 | I can laugh and see the funny side of things | 0.279 | 0.633 |
| D12 | I look forward with enjoyment to things | 0.248 | 0.574 |
| D6 | I feel cheerful | 0.380 | 0.563 |
| D10 | I have lost interest in my appearance | 0.345 | 0.431 |
| D8 | I feel as if I am slowed down | 0.400 | 0.425 |
| Eigenvalues | 5.67 | 1.40 | |
| Variance (%) | 40.55 | 10.03 | |
| Cronbachs alpha | 0.82 | 0.83 | |
HADS: Hospital anxiety and depression scale.
Adequate congruence coefficients were obtained for the anxiety factor (0.95, p=0.001) and depression factor (0.91, p=0.001).
Discussion
Given the high prevalence and comorbidity of anxiety, depression, and CVD2 and the negative consequences in health and quality of life carried out by their simultaneous presentation8,9, there is a need for a valid, reliable, and easy to apply measure to assess and then treat such emotional disorders in the setting of medical practice. The HADS was specially designed to fulfill this purpose14. Thus, the aim of the present study was to evaluate the basic psychometric properties of the HADS Spanish version in Mexican patients with a confirmed diagnosis of ischemic or hypertensive cardiomyopathy. According to our data, the Spanish version of HADS is characterized by construct validity and reliability in Mexican patients with CVD, what is reflected in its factorial structure and internal consistency coefficients, respectively.
First, in our sample, a bifactor structure integrating all items for anxiety and depression in two different scales allows the identification and differentiation of both emotional disorders as in the original version of the instrument and in contrast with the one single factor found by Herrera, Galindo, and Lerma22 in a smaller and less controlled (for cognitive deficit or psychiatry disorder presence) sample of Mexican patients with CVD.
This is important because, although anxiety and depression coexist in many patients, there are patients who only require treatment for one of such conditions, and at least psychological interventions for each emotional problem are quite different. Thus, a measure that identifies and differentiates the presence and severity of both conditions is much more clinically useful in terms of prescription of needed treatment in a more parsimonious way. For example, in the case of evidence-based behavioral intervention, patients with only anxiety will require relaxation training27 that is not mandatory for those with only depression, which must need behavioral activation28 (that is not the main strategy to treat anxiety). The screening capacity of the instrument for depression and anxiety is a valuable tool for most of the public health institutions in Mexico, where human and material resources for mental health are limited.
In addition, according to Nunnallys suggestion for the interpretation of Cronbachs alpha coefficients (to consider them high as from 0.70)24, the internal consistency indexes were adequate for both anxiety (0.82) and depression (0.83) scales. These findings are consistent with those obtained from the original version of the instrument (0.83 and 0.80, respectively)1, in different countries, including Mexico, and diverse groups of patients2,15-20, showing a bi-factorial structure of the HADS.16-20. In addition, the high coefficients obtained in the congruence analysis give further support of the adequacy of the scale for its use in patients from medical settings.
Conclusion
The present study adds evidence of the validity and reliability of the HADS to assess the presence of both depression and anxiety in Mexican patients with chronic medical diseases, in general, and with CVD, in particular. Its central contribution, given the previous studies in Mexican HIV patients and CVD samples that showed unifactorial structures that suggest the measurement of distress in general, is that it demonstrates the capacity of the measure to also differentiate the presence and severity of the two emotional problems most frequently in medically ill population, depression, and anxiety. This makes the HADS an adequate tool to understand the specific experience of suffering in the setting of medical practice, and the explicit intervention needed to reduce it.
