A mixed methods case study exploring the impact of membership of a multi-activity, multicentre community group on social wellbeing of older adults

Table 2 Means and standard errors for social wellbeing variables over time

Variable	Baseline (n = 28)	6 months (n = 27)	12 months (n = 28)	p-value
Duke social support^a	3.83 (0.2)	4.08 (0.2)	4.09 (0.2)	0.065^d
DJG loneliness^b	27.95 (1.94)	26.18 (1.94)	25.17 (1.94)	0.028*
UCLA loneliness^c	5.31 (0.39)	4.64 (0.39)	4.93 (0.39)	0.023*^d

*significant effect of time for the indicated variable at p < 0.05
All analyses are adjusted for age, employment at baseline, and mean weekly LAC attendance
^aDuke_UNC functional social support scale. Range 1–5. High social support = 5. p-value represents the p-value for the log-transformed variable
^bDe Jong Gierveld loneliness scale. Scored as a 5 item likert scale from Yes!, yes, more or less, no, No! Range = 11–55. Highest loneliness = 55.
^cUCLA 3-item loneliness scale. Range = 3–9. Highest loneliness = 9
^dp-value presented here are log-transformed variable analyses

ISSN: 1471-2318