^{1}

^{2}

Depressive symptoms among university students are a major mental-health issue worldwide, and university students are particularly vulnerable to various stressors that can produce depression. Therefore, accurate and sustainable assessment of depressive symptoms among university students is of special importance. The Patient Health Questionnaire-9 (PHQ-9) is one such measure. The objective of the current study was to evaluate the psychometric properties of the PHQ-9 among Korean university students.

A total sample of 582 university students attending a four-year private university in South Korea was recruited for the study. Confirmatory factor analyses (CFAs) were performed to compare the goodness-of-fit of four competing models suggested by extant literature on the PHQ-9. Convergent validity was assessed using a correlation analysis between the PHQ-9 and other psychiatric instruments, including the Depression, Anxiety, and Stress Scale-21 (DASS-21) and the Generalized Anxiety Disorder Scale (GAD-7).

A one-factor structure of the PHQ-9 provided the best fit to the data. Internal consistency was adequate. The PHQ-9 demonstrated good convergent validity with related constructs.

The psychometric properties of the PHQ-9 proved to be adequate, with a robust and interpretable factor structure and good internal consistency. The PHQ-9’s validity, reliability, brevity and ease of administration make it a useful screening instrument for depression among university students in Korea.

Depression (i.e., major depressive disorder or clinical depression) is the most common psychiatric disorder in the general population and is characterized by persistent feelings of sadness, loss of interest or pleasure, neurovegetative disturbance, and reduced energy [

Entering university can be a highly stressful life transition for many or even most students, bringing with it increased exposure to stressors. During this crucial stage, students commonly experience financial stress, academic demands, fear of failure or future, and stress in interpersonal relationships [

The psychometric properties of the PHQ-9 have been evaluated with various population, including psychiatric patients [

Although the PHQ-9 has been well validated in primary care settings [

Data were collected from 582 undergraduate students (234 male and 348 female) attending a four-year private university in the central region of Korea. The students’ courses of study included architecture, design, education, and social and child welfare. The participants ranged in age from 18 to 30 years (M=20.3, SD=2.00), with the great majority (70%) between the ages of 19 and 23. The mean age was 19.8 for female students (SD=1.51) and 21.0 for male students (SD=2.39).

This study was approved by the Institutional Review Board of Woosong University (Protocol Code: 1041549-190709-SB-76). We made arrangements with academic instructors for students to complete the questionnaires as a group during scheduled class time. Either the principal investigator or the coauthor was present at each administration to provide instructions. The questionnaire was administered in paper-and-pencil form. Written informed consent was obtained from all participants after the aims and objectives of the study had been explained. Participants were informed about the content of the questionnaires and that participation was entirely voluntary. Participants were also assured that the anonymity and confidentiality of their responses would be protected. On average, the questionnaire took about 15 to 20 minutes to complete.

The PHQ-9 [

The Depression Anxiety and Stress Scale-21 (DASS-21) [

The Generalized Anxiety Disorder-7 (GAD-7) [

The data were analyzed using IBM SPSS Statistics for Windows, Version 23.0 (IBM Corp., Armonk, NY, USA) and AMOS 20.0 (IBM Corp.) [

Next, CFA with the maximum-likelihood procedure was used to test a series of alternative plausible models for the structure of the PHQ-9. CFA is used to test statistically whether a hypothesized linkage pattern between the observed variables and their underlying latent constructs actually exists [^{2}) and its subsequent degrees of freedom (df); CFI; goodness-of-fit index (GFI); root mean square error of approximation (RMSEA) and its 90% confidence interval (90% CI); and standardized root mean square residual (SRMR). CFA values of 0.95 or higher indicate good fit [

Four competing models suggested by the extant literature were tested and the resulting fit indices were compared to assess how well each model fit the data. Model 1 is the original one-factor model hypothesized by Kroenke et al. [

After we examined the goodness-of-fit with the PHQ-9 for competing models, we used chi-square difference tests to determine whether the models were significantly different from one another. To assess the convergent validity of the PHQ-9, the associations between the PHQ-9 and the criteria instruments DASS-21 and GAD-7 were examined using correlation with Pearson’s r. Internal consistency was determined by calculation of Cronbach’s alpha.

The participants’ scores ranged from 0 to 27, with a mean of 5.72 (SD=0.28). Hence, the current sample overall tended to exhibit only mild symptoms of depression. With regard to reliability, the internal consistency of the PHQ-9 in this sample indicated an adequate Cronbach’s alpha of 0.83.

^{2}=223.9, df=27; CFI=0.91; GFI=0.98; RMSEA= 0.089 (90% CI=0.074–0.085); SRMR=0.051] (^{2}=1069.8, df=26; CFI=0.53; GFI=0.84; RMSEA=0.263 (90% CI=0.250–0.276); SRMR=0.225]. Model 2b fit the data marginally better than Model 2a, as evidenced by a slight decrease in the chi-square value, CFI, GFI, RMSEA, and SRMR, but the indices did not meet the accepted fit criteria [^{2}=1065.3, df=26; CFI=0.57; GFI=0.87; RMSEA=0.232 (90% CI=0.249–0.135); SRMR=0.224]. Finally, Model 2c yielded almost identical fit indices to Model 2a [^{2}=1069.1, df=26; CFI=0.53; GFI=0.84; RMSEA=0.263 (90% CI=0.249–0.276); SRMR=0.225]. More importantly, this lack of fit is further underscored by a significant chi-square difference between Model 1 and each of Models 2a [^{2} (1)=845.8, p<0.001), 2b (^{2} (1)=841.4, p<0.001], and 2c [^{2} (1)=845.2, p<0.001].

Thus, the one-factor model (Model 1) was deemed the best fit to the data for the entire sample, which appeared to be unidimensional rather than multidimensional. All standardized factor loadings for the one-factor model were significant at p<0.01 and ranged from 0.53 to 0.79, indicating good factor loading.

The PHQ-9 was moderately and positively correlated with scores on the three subscales of the DASS-21, including depression (r=0.69), anxiety (r=0.60), and stress (r=0.62), as well as with the GAD (r=0.68), which measures anxiety. Convergent validity was, therefore, satisfied for the PHQ-9 with our sample.

To the best of our knowledge, this is the first study to examine the psychometric properties of the PHQ-9 among Korean university students. Testing of the original version of the PHQ-9 with a large primary care sample in a previous study revealed that the instrument has strong psychometric properties, as evidenced by its good reliability and good criterion, factorial, and convergent validity [

Additionally, Granillo [

Given that depression among university students is a major concern in university mental-health settings, our results have important implications for the assessment and diagnosis of depressive symptoms among Korean university students. As Keum et al. [

Several limitations of this study should be considered. First, the generalizability of the current findings may be limited because our study sample was relatively homogeneous (consisting solely of university students). Therefore, the factorial structure of the PHQ-9 may be different for various populations such as adolescents, older adults, individuals with more severe clinical conditions, and culturally diverse populations. Although the current study supported a one-factor model of the PHQ-9 as a valid measure for Korean university students, this finding does not mean that the multidimensional model of the PHQ-9 is invalid. It is possible that the competing two-factor models may be more appropriate with other populations. Replication with more heterogeneous samples is thus needed to examine the utility of the PHQ-9 across diverse populations [

Within these limitations, our findings indicate that the PHQ-9 is a psychometrically sound and unidimensional depression measure for Korean university students. Our results add to the evidence that the PHQ-9 is an efficient and valid self-report depression measure for subjects in both clinical and non-clinical settings.

The authors have no potential conflicts of interest to disclose.

Conceptualization: Boram Lee. Data curation: Yang Eun Kim. Formal analysis: Boram Lee. Funding acquisition: Yang Eun Kim. Investigation: Boram Lee. Methodology: Boram Lee. Project administration: Yang Eun Kim. Resources: Yang Eun Kim. Software: Boram Lee. Supervision: Boram Lee. Validation: Boram Lee. Visualization: Yang Eun Kim. Writing—original draft: Boram Lee, Yang Eun Kim. Writing—review & editing: Boram Lee, Yang Eun Kim.

One-factor model of the PHQ-9 [N=582; ^{2} =223.9; df=27; CFI=0.91; GFI=0.98; RMSEA=0.089 (90% CI=0.076–0.098); SRMR=0.051]. PHQ-9: Patient Health Questionnaire-9, df: degrees of freedom, CFI: comparative fit index, GFI: goodness of fit index, RMSEA: root mean square error of approximation, SRMR: standardized root mean residual.

Sociodemographic information of participants

Sociodemographic information | Frequency | Percentage (%) |
---|---|---|

Gender | ||

Male | 234 | 40.2 |

Female | 348 | 59.8 |

Age in years | ||

18 | 125 | 21.5 |

19–23 | 407 | 69.9 |

24–29 | 49 | 8.4 |

30 | 1 | 0.2 |

Mean (SD) | 20.3 (2.00) | |

Degree of subject | ||

Architecture | 176 | 30.2 |

Design | 150 | 25.8 |

Education | 64 | 11.0 |

Social and child welfare | 192 | 33.0 |

Goodness-of-fit indices of models for the PHQ-9 (N=582)

Model | κ | χ^{2} |
df | CFI | GFI | RMSEA (90% CI) | SRMR |
---|---|---|---|---|---|---|---|

Model 1^{†} |
9 | 223.9^{*} |
27 | 0.91 | 0.98 | 0.089 (0.076–0.098) | 0.051 |

Model 2a | 9 | 1069.8^{*} |
26 | 0.53 | 0.84 | 0.263 (0.250–0.276) | 0.225 |

Model 2b | 9 | 1065.3^{*} |
26 | 0.57 | 0.87 | 0.232 (0.249–0.135) | 0.224 |

Model 3c | 9 | 1069.1^{*} |
26 | 0.53 | 0.84 | 0.263 (0.249–0.276) | 0.225 |

p<0.01,

represents a final model used in the study.

PHQ-9: Patient Health Questionnaire-9, K: number of items, df: degrees of freedom, CFI: comparative fit index, GFI: goodness of fit index, RMSEA: root mean square error of approximation, SRMR: standardized root mean residual

Standardized factor loadings for the modified model of the PHQ-9

PHQ-9 item | Factor loadings |
---|---|

1. Anhedonia | 0.73 |

2. Depressed mood | 0.79 |

3. Sleep difficulties | 0.61 |

4. Fatigue | 0.53 |

5. Appetite changes | 0.54 |

6. Feeling of worthlessness | 0.53 |

7. Concentration difficulties | 0.74 |

8. Psychomotor agitation/retardation | 0.70 |

9. Thoughts of death | 0.72 |

PHQ-9: Patient Health Questionnaire-9