To validate and translate the English version of the Neurologic Depression Disorders Inventory in Epilepsy (NDDI-E) into Spanish as a screening instrument for major depressive episodes (MDE) for patients with epilepsy from Argentina and Uruguay.
One hundred fifty-five consecutive outpatients with epilepsy participated in this study. The module of MDE of the MINI International Neuropsychiatric Instrument (MINI Plus version) was used as the gold standard against which the translated version of the NDDI-E was validated.
Among the 155 patients, 25 (16%) met Diagnostic and Statistical Manual, Fourth Edition (DSM-IV) criteria for MDE according to the MINI. With a total score of >15, The NDDI-E identified MDE with an 80% sensitivity, 90% specificity, 60% positive predictive value, and 95.5% negative predictive value.
These data indicate that the Spanish version of the NDDI-E can reliably identify MDE in patients with epilepsy from Argentina and Uruguay.
Depression is the most frequent psychiatric comorbidity in patients with epilepsy (PWE),1 with lifetime prevalence rates estimated to range between 30% and 35%.2 Major depressive episodes (MDEs) are the most severe forms of depressive disorders. They are associated with a worse quality of life,3 increased suicidal risk,4 worse tolerance to antiepileptic drugs (AEDs),5 worse response to pharmacologic and surgical treatment of the seizure disorder,6,7 and an increased burden on the patient, family, and society as a whole.8,9 Yet, despite the relatively high prevalence of MDE and its negative impact on the life of PWE, it remains under-recognized and undertreated.10–12
Screening of MDE with self-rating instruments may be a potential solution for the recognition of this psychiatric comorbidity. The Neurologic Depressive Disorders Inventory in Epilepsy is a six-item, self-rating instrument developed in the United States to identify MDE specifically in PWE. It takes 3 min to complete; a total score >15 is suggestive of a MDE with a sensitivity of 89% and a specificity of 90%.13 Since its publication in 2006, there has been an interest in validating and translating this instrument to other languages, and up to now, validated versions of the NDDI-E have become available in Portuguese,14 Japanese,15 Italian,16 and Spanish from Spain.17 In this article, we present the validation and translation of the NDDI-E from the original English version into Spanish from Argentina.
Translating instruments or questionnaires is a difficult but cost-effective process that provides the tools for evaluating the same issue in separate communities.18,19 The need to develop an NDDI-E version for Argentina was prompted by the significant differences in the vocabulary, idioms, and local expressions between the Spanish language in Spain and Argentina, which would limit the validity of the Spanish version from Spain in PWE from Argentina.
Between September and November 2010, a total of 155 outpatients with epilepsy (age 16 years old and older) were recruited from the epilepsy clinics of the Institute of Neurosciences at the Favaloro Foundation and the British Hospital in Buenos Aires, Argentina.
We used the MINI, against which we measured our NDDI-E results. The Mini International Neuropsychiatric Interview (MINI) is a previously validated interviewer-administered, structured, diagnostic, psychiatric interview that renders a dichotomous classification of major psychiatric disorders.
Although we are aware that the MINI and the NDDI-E were intended for individuals oldder than the age of 18, as it is a consecutive sample, we included one 16-year-old and one 17-year-old. We know this could be a limitation for this study. Every patient had to have a diagnosis of focal or primary generalized epilepsy, documented clinically and electrographically, and had to be on one or two antiepileptic drugs (AEDs). They had to have completed at least primary school (6 years) and be able to read fluently and understand the questionnaires and rating scales. Patients unable or unwilling to complete or understand the study protocol and consent forms, with psychotic symptoms or a dementing process at the time of enrollment, were excluded from the study. This is a sample of consecutive patients evaluated in our hospital, so the severity of the epilepsy is variable, due to the random sample. We did not examine seizure severity because our objective was the validation of the scale in people with epilepsy regardless of the severity of their seizures.
Demographic data included age, gender, marital status, highest grade achieved in school/college, and professional status. Epilepsy-related data included type of epilepsy syndrome, type of seizures, number and type of AEDs, and neuroimaging and electroencephalographic data (obtained from routine electroencephalography [EEG] studies and/or video-EEG monitoring studies).
The presence of MDE was established with the MDE module of the MINI International Neuropsychiatric Instrument, which is a structured interview that identifies current and past MDE.20 This instrument was used as the gold standard measure against which the validation of the NDDI-E was conducted. In addition to the MDE module, patients were given the other four modules of mood disorders (modules A, A5-6, B, and C).
Translation of the NDDI-E to Spanish
The original English version of the NDDI-E13 was first translated into Spanish by one group of official translators. This first Spanish version was then translated back into English by an independent group, and the version was sent to one of the authors of the original English version (AMK) for approval.
The presence of MDE was investigated in every patient with the MINI. The total score on the NDDI-E, Spanish version was calculated, and these data were entered into a data base using IBM SPSS Statistics version 19.0 (Armonk, New York, U.S.A.). Proportions were calculated for categorical variables. Pearson chi-square and Fisher's exact test were used for comparison between the group of PWE with and without MDE. Continuous variables between the two groups were compared using nonparametric Mann-Whitney U-tests. Cronbach's coefficient α, and item−total (corrected item−total correlation) and interitem correlations (Spearman's ρ) were computed to ascertain the internal consistency of the Spanish version of the NDDI-E and receiver operating characteristic (ROC) curve analyses was performed.
Among the 155 PWE enrolled, 59 (38%) were male. The average age was 47 ± 18 years. Every subject had completed at least a high school education. No patient had difficulty completing the Spanish version of the NDDI-E. The demographic, epilepsy, and psychiatry data appear on Table 1. Thirty-four patients were taking antidepressant medication: 26 (20.6%) among nondepressed patients and 8 (32%) among the depressed patients. All patients who were taking antidepressant medication had been doing so for <1 year.
Table 1. Demographic, psychiatric, and epilepsy-related variables
Nondepressed (n = 130)
Major depression (n = 25)
Mean age, years (SD)
Seizure type (%)
Drug-resistant epilepsy (%)
Currently taking antidepressive medication (%)
Mean NDDI-E (SD)
One hundred twenty patients (77%) had focal epilepsy and 33 (21%) had generalized epilepsy; 129 patients (83%) were taking one AED. Patients with and without MDE did not differ in any of the epilepsy-related variables (Table 1).
Psychometric properties of NDDI-E
Cronbach's alpha for the NDDI-E was 0.77, which represents an acceptable internal consistency reliability. All NDDI-E items were positively associated with the total score. Receiver operating characteristic (ROC) curve analysis of the NDDI-E showed an area under the curve of 0.905 (95% CI 0.845–0.965; see Fig. 1). At a cutoff score of >15, the NDDI had a sensitivity of 80%, a specificity of 90%, a positive predictive value (PPV) of 60.6%, and a negative predictive value (NPV) of 95.9%.
Twenty-five patients (16%) met diagnostic criteria of MDE with the MINI. Using a cutoff score of >15, the NDDI-E identified correctly 20 (80%) of these patients. In addition, a total of 13 patients (8.4%) had a total score >15, but did not meet diagnostic criteria of MDE according to the MINI. As expected, patients who met diagnostic criteria of MDE had significantly higher total scores on the NDDI-E (17.2 ± 3.27) than nondepressed patients (10.65 ± 3.58; p < 0.001).
These data suggest that the Spanish version (for Argentina) can be used as a valid instrument to screen for MDE in PWE in Spanish-speaking patients from Argentina. This instrument can be also used in Uruguay, where the language is identical. The translated version of the NDDI-E has good psychometric characteristics that are comparable to those of the original English version,13 including its specificity and sensitivity, with the use of the same total cutoff score (>15).
This is the second version of the NDDI-E that has been validated and translated into Spanish; the first one was developed for Spain.17 As stated previously, because of the differences in the Spanish spoken in Spain and Latin American countries and in particular, Argentina and Uruguay, a separate version of the NDDI-E was needed for these two countries. The number of patients enrolled and protocols followed by the Spanish and Argentinean investigators were very similar. Table 2 shows a comparison of the psychometric data of the two Spanish versions. The main differences between the two instruments included a lower total cutoff score (>13 vs. >15), and a lower specificity (78% vs. 90%) in the version from Spain.
Table 2. Psychometric properties of the Spanish NDDI-E versions from Argentina and Spain
Corrected item-total correlation
Cronbach's alpha if item deleted
Corrected item-total correlation
Cronbach's alpha if item deleted
Everything is a struggle
Nothing I do is right
I'd be better off dead
Difficulty finding pleasure
Table 3 shows the two Spanish versions. In three of their six items, the version from Spain included a verb that was not used in the Argentinean version, illustrating the differences in the language used in each country. Our version yielded the same cutoff total score identified in the original English version,13 the Portuguese version for Brazil,14 and the Japanese version.15 Of note, there were no differences in the demographic and seizure-related variables between the Spanish and Argentinean cohorts. The authors of the Spanish version endorsed less frequently suicidal ideation (item: I'd be better off dead) than the Argentinean subjects (25% vs. 75%). This item may account for the differences in the total score of the two versions.
Table 3. Spanish versions of the NDDI-E from Argentina and Spain
Todo es una lucha
Todo me supone un esfuerzo
Nada de lo que hago sale bien
Nada de lo que hago me sale bien
Me siento culpable
Me siento culpable
Estaría mejor muerto
Siento que estaría mejor muerto
Me siento frustrado
Dificultad para encontrar placer
Tengo dificultad para sentir placer
In conclusion, the Argentinean version of the NDDI-E appears to be a valid instrument to screen for MDE in PWE. The differences of the total cutoff scores between the two available Spanish versions suggest the need to test this instrument in other Spanish-speaking countries. The idioms and different style of Spanish spoken in Argentina and Uruguay compared to the rest of Latin America provide the rationale for the validation of this Spanish version, which would need to be revalidated if it were to be used in other Spanish-speaking countries. The need for multiple versions is in part a limitation of these instruments, but provides the advantage of a reliable tailored version adapted specifically to the linguistic characteristics of each country.
Dr. Alfredo Thomson has received honoraria for lectures given on behalf of the following pharmaceutical companies: Abbott, Novartis, GlaxoSmithKline, Pfizer, UCB Pharma, Jansen-Cilag, Gador Laboratory, and Raffo Laboratory. He has also obtained honoraria for participation in clinical trials from the following pharmaceutical companies: Novartis, Jansen-Cilag, Eisai, and Upsher-Smith. Dr. Andres Kanner declares that he received an honorarium for services rendered to Neuropace and Vertex Laboratories as a consultant. Drs. Analia Calle, Elena M Fontela ME, Luis Yepez, Francisco Muñoz Giacomelli, Agustin Jáuregui, and Joaquin M Racosta have no disclosures to report. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Dr. Alfredo E. Thomson is involved in clinical research on psychiatric aspects of epilepsy at the Clinica Favaloro in Buenos Aires, Argentina.