Abstract: Objective To evaluate the psychometric properties and applicability of the 6-item University of California Los Angeles (UCLA) Loneliness Scale (ULS-6) in adults. Methods We conducted 2 surveys to assess the validity of different measurement scales and questionnaires. In Survey 1, a total of 1480 adults were measured using the UCLA Loneliness Scale (ULS), Patient Health Questionnaire- 9 (PHQ-9) and Perceived Social Support Scale (PSSS), and the data were used for item analysis and assessment of the reliability, validity and measurement invariance. In Survey 2, UCLA Loneliness Scale was used for measurement in 652 college students, and the data were used for analysis of the criterion validity of ULS-6; 3 weeks later, 300 of the students were retested using ULS-6 to assess the retest reliability of the scale. Results Item analysis suggested that the items in ULS-6 all had good discrimination power with discrimination indexes all above 0.775 (r=0.775-0.820, P<0.001). Measuring only one dimension, ULS-6 had an internal consistency reliability of 0.891, a split-half reliability of 0.875, and a retest reliability of 0.726. The correlation coefficients of ULS-6 with ULS, ULS-8, PHQ-9 and PSSS were 0.882, 0.967, 0.528 and -0.532, respectively. The measurement invariances of ULS-6 across genders and age groups were all acceptable. Among the adult participants, the mean total score of ULS-6 was 12.97±3.96; While only 20% of the adults had no loneliness, 80% of them exhibited varying degrees of loneliness, ranging from mild (39.6% ) and moderate (25.7% ) to intense (14.7% ) feelings of loneliness. Conclusion The ULS-6 has good reliability, validity and applicability for measurement of loneliness in Chinese adults.

Key words: 6-item University of California Los Angeles Loneliness Scale; loneliness; reliability; validity; measurement invariance