The factor structure of the Pittsburgh Sleep Quality Index in persons with traumatic brain injury. A NIDRR TBI model systems module study

Abstract

BACKGROUND: Sleep disturbances are common following traumatic brain injury (TBI). The Pittsburgh Sleep Quality Index (PSQI) is a widely used measure of sleep quality that has been used in numerous populations. Although this measure has been used in TBI research, there are few studies examining the psychometric properties in this population. OBJECTIVE: The current study examined the factor structure of the PSQI in a sample of persons with TBI and tested the one, two, and three factor models derived from previous studies in other populations. METHODS: A telephone interview was conducted with 243 individuals who had sustained a TBI. All participants were approximately one year post-injury. Factor analyses were conducted (exploratory and confirmatory) to examine the factor structure of the PSQI. RESULTS: Results confirm the fit of models previously tested in the literature but also reveal an alternative conceptualization of sleep containing both qualitative and quantitative factors. CONCLUSIONS: While the 3-factor model best fits the data in this TBI sample, the use of a 2-factor model is acceptable and may be more clinically relevant due to the grouping of time-related variables that could provide important information with regard to circadian rhythm disorders.

Dimensionality of the Pittsburgh Sleep Quality Index: a systematic review

“…The factor analysis of the PSQI has been reported on diverse samples including university/college students in Nigeria, India, Chili, Chinese, Ethiopia, Peru and Thailand, pregnant women, community dwelling adults and older adults in America, Australia, China, Ethiopia, Spain, Portugal, Iranian health professionals, adolescents and young adults [ 15 – 17 , 23 , 24 , 26 – 28 , 31 , 34 , 36 , 39 , 41 , 44 , 45 , 52 , 53 , 57 – 60 ]. Moreover, it has been reported for patient population with breast cancer, Crohn’s disease, depression, diabetes, women with hot flashes, Taiwanese insomniacs, post-traumatic stress disorder, Trauma exposed veterans, temporo-mandibular disorder, traumatic brain injury, fibromyalgia, arthritis, chronic fatigue syndrome, psychiatric disorders and renal transplant patients [ 25 , 27 , 29 , 30 , 32 , 33 , 35 , 37 , 38 , 40 , 42 , 43 , 46 , 49 , 51 , 61 – 65 ].…”

“…We counted the number of studies reporting any of the following 5 indices: Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy, Bartlett’s sphericity test, determinant score, anti-image, and component correlations (Table 1 ). Eighteen studies reported both the KMO and Bartlett’s sphericity test [ 15 – 17 , 25 , 39 , 43 – 45 , 51 – 53 , 57 – 59 , 61 , 64 , 65 , 67 ], reported inter-item correlations [ 16 , 22 , 24 , 25 , 27 , 29 , 30 , 32 – 34 , 36 , 37 , 41 , 42 , 45 , 46 , 50 , 53 , 57 , 64 – 67 ], and 11 studies did not report any index [ 23 , 26 , 28 , 31 , 35 , 38 , 40 , 49 , 60 , 62 , 63 ].…”

“…Twenty articles did not report having carried out EFA [ 15 , 23 , 26 , 29 , 32 – 38 , 42 , 48 , 50 , 59 , 60 , 62 – 64 , 66 ]. The types of EFA tests used were principal component analysis (PCA) in 18 studies [ 17 , 22 , 24 , 25 , 27 , 28 , 31 , 39 , 41 , 43 – 45 , 49 , 51 , 57 , 58 , 65 , 67 , 68 ], maximum likelihood estimation (MLE) in 6 [ 16 , 17 , 22 , 24 , 40 , 41 , 53 ] and unreported in 2 [ 30 , 52 ] (Table 2 ). The types of rotation used for EFA were orthogonal varimax in 8 studies [ 28 , 39 , 41 , 43 , 44 , 46 , 49 , 52 , 65 ], oblique direct oblimin in 8 [ 16 , 17 , 22 , 24 , 25 , 51 , 53 , 61 ], oblique promax in 5 [ 40 , 45 , 46 , 58 , 65 ], standardized geomin 1 [ 30 ], and unreported...…”

“…Seven studies reported using a Cattell’s scree test to determine the number of factors to retain [ 16 , 17 , 44 , 46 , 58 , 61 , 65 , 69 ]. Fifteen studies employed the Kaiser criterion of eigenvalue greater than one [ 16 , 17 , 25 , 27 , 28 , 40 , 43 – 46 , 49 , 51 , 58 , 61 , 65 ].…”

“…However, only 4 studies used parallel analysis [ 16 , 17 , 49 , 65 , 71 ]. Fifteen studies used multiple criteria for factor retention [ 16 , 17 , 25 , 27 , 40 , 43 – 46 , 49 , 51 , 52 , 58 , 61 , 65 ]. The number of factors retained after EFA varied across articles.…”