Objectives: First, to seek if sociodemographic and health factors contribute differentially to the explanation of loneliness in institutionalized and noninstitutionalized older adults; and second, to analyze the influence of institutionalization on loneliness. Method: This work was based on two surveys of older adults aged 60 years or more in Spain. A group of 234 community-dwelling people and 234 nursing homes residents were selected (n = 468). Logistic regression models were applied using the six-item De Jong Gierveld Loneliness Scale as dependent variable. Results: Depression was associated with loneliness in both populations. Sex and marital status contributed to explain loneliness among those living at home, whereas gathering with family, friends, and neighbors showed a significant effect in the institutionalized group. Institutionalization per se showed a strong effect on loneliness. Discussion: Findings have potential implications for targeting older adults at risk for loneliness.

Aim of this study was to validate the Italian version of the Brief Pain Inventory (BPI), Breve Questionario per la Valutazione del Dolore (BQVD), which is a multidimensional instrument to assess pain intensity and pain interference with daily functions. A group of 110 patients with cancer pain were enrolled in the study and were administered the BQVD and the Therapy Impact Questionnaire (TIQ) which is a valid instrument for quality of life assessment in cancer patients. Cronbach's alphas were computed for the interference and severity scales in assessing reliability. Confirmatory factor analysis was utilized to ascertain construct validity of the BQVD. Measures of interference and of intensity were calculated a priori from the TIQ result and were used in correlated correlations. Alpha coefficients for the pain severity and the pain interference scale were above 0.75. Confirmatory factor analysis showed that a 2-factor solution for the BQVD structure was interpretable and provided adequate fit for the data. The correlation with the TIQ items showed a stronger association between factor 1 (interference) and the interference with affect and activity measure from the TIQ, while factor 2 (severity) was more strongly associated with the TIQ pain severity measure. In comparison with other non-italian samples our results show a lower reliability estimate. Overall the analysis of these data shows that the BQVD is a useful and valid tool in assessing pain and its impact on patients' quality of life which could also help in developing international and cross-cultural studies in cancer pain.