The association between specific activity components and depression in nursing home residents: the importance of the social component

Figures & data

Table 1. Examples of activities with their corresponding component rank scores (levels: A = absent or very low, L = low, M = medium, H = high).

Activities	Physical	Creative	Social	Cognitive	Musical
Reading in their own room (book, magazine, newspaper)	A	A	A	M	A
Discussing news or sports in a group	A	A	H+	M++	A
Exercise alone in their own room by a workout program on television	H++	A	A	L+	A+
Walking outside in a group (with or without using a rollator)	M++	A	M	A+	A
Handcrafts alone in their own room (knitting, embroidering)	A++	M+	A	L+	A
Painting alone in their own room	A+	H++	A	M	A
Painting in a group	A+	H++	L++	M	A
Memory exercises in a group	A	A+	H	H++	A
Memory exercises alone in their own room by a program on television	A	A+	A	H++	A
Listening to music alone in their own room	A	A	A	A+	M+
Playing the keyboard alone in their own room	A++	M+	A	M++	H++

Note. ‘+’ and ‘++’ reflect intermediate rank scores that were added when the activity was ranked to a higher rank by one or two researchers, or when discrepancies regarding equivalence or distinctiveness were solved by re-ordering the activities on the component.

Table 2. Socio-demographic factors, depressive symptoms, and antidepressant use.

	t0 (N, 40)	t1 (N, 33)	t2 (N, 31)
Age, mean (SD) [range]	79.8 (8.9) [60–96]	80.1 (9.4) [60–96]	79.5 (9.4) [60–96]
Sex, female, N (%)/male, N	26 (65)/14	23 (69.7)/10	21 (67.7)/10
Marital status
Unmarried, N (%)	2 (5)	2 (6.1)	2 (6.5)
Married or partnered in a registered partnership, N (%)	8 (20)	6 (18.2)	6 (19.4)
Widowed (after marriage or registered partnership), N (%)	26 (65)	21 (63.6)	19 (61.3)
Divorced (after marriage or registered partnership), N (%)	4 (10)	4 (12.1)	4 (12.9)
Educational attainment
Low, N (%)	25 (62.5)	20 (60.6)	19 (61.3)
Medium, N (%)	14 (35)	12 (36.4)	11 (35.5)
High, N (%)	1 (2.5)	1 (3)	1 (3.2)
Depressive symptoms, mean (SD) [range]	19.6 (13.4) [1–46]	18.9 (13.3) [1–50]	19.6 (13.1) [2–51]
Antidepressant use, yes, N (%)/no, N^a	15 (37.5)/25	12 (36.4)/21	10 (32.3)/21

Note. ^aMissing values (N_t0 = 5, N_t1 = 3, N_t2 = 2) were replaced by antidepressant use is ‘no’.

Table 3. Number of nursing home residents per activity component level, N (%).

	t0 (N, 40)	t1 (N, 33)	t2 (N, 31)
Physical
Absent	16 (40)	11 (33.3)	11 (35.5)
Low	8 (20)	7 (21.2)	6 (19.4)
Medium	8 (20)	5 (15.2)	9 (29.0)
High	8 (20)	10 (30.3)	5 (16.1)
Creative
Absent	18 (45)	16 (48.5)	13 (41.9)
Low	7 (17.5)	8 (24.2)	9 (29)
Medium	7 (17.5)	4 (12.1)	5 (16.1)
High	8 (20)	5 (15.2)	4 (12.9)
Social
Absent	4 (10)	3 (9.1)	6 (19.4)
Low	12 (30)	7 (21.2)	6 (19.4)
Medium	12 (30)	7 (21.2)	7 (22.6)
High	12 (30)	16 (48.5)	12 (38.7)
Cognitive
Absent	0 (0)	0 (0)	0 (0)
Low	13 (32.5)	14 (42.4)	16 (51.6)
Medium	13 (32.5)	6 (18.2)	6 (19.4)
High	14 (35)	13 (39.4)	9 (29)
Musical
Absent	23 (57.5)	18 (54.5)	17 (54.8)
Low	5 (12.5)	8 (24.2)	6 (19.4)
Medium	6 (15)	1 (3)	4 (12.9)
High	6 (15)	6 (18.2)	4 (12.9)

Note. For each resident, five total component scores were calculated and categorized using cut-off scores determined at baseline into levels: absent (scores of 0), low (first 33.3% of scores higher than 0), medium (second 33.3% of scores higher than 0), and high (last 33.3% of scores higher than 0).

Table 4. Mixed models of activity components predicting depressive symptoms.

Parameter	Physical component	Creative component	Social component	Cognitive component	Musical component	All activity components
AD, yes^a	0.02 (0.22)	−0.06 (0.21)	0.05 (0.19)	−0.15 (0.21)	−0.10 (0.22)	0.07 (0.21)
Activity components
Physical^b	F(3, 53.61) = 2.73, p = .053					F(3, 54.55) = 0.29, p = .831
Low	−0.23 (0.13)					0.13 (0.19)
Medium	−0.23 (0.17)					0.15 (0.21)
High	−0.47 (0.17)					−0.00 (0.23)
Creative^b		F(3, 52.44) = 2.52, p = .068				F(3, 51.11) = 1.16, p = .336
Low		−0.19 (0.14)				−0.25 (0.18)
Medium		−0.44 (0.17)				−0.42 (0.23)
High		−0.35 (0.21)				−0.17 (0.26)
Social^b			F(3, 57.73) = 7.92, p < .001			F(3, 54.96) = 4.73, p = .005
Low			−0.39 (0.20)			−0.36 (0.25)
Medium			−0.82 (0.21)			−0.78 (0.27)
High			−0.82 (0.20)			−0.75 (0.27)
Cognitive^c				F(2, 62.21) = 4.23, p = .019		F(2, 65.52) = 1.02, p = .368
Medium				−0.37 (0.15)		−0.21 (0.16)
High				−0.40 (0.15)		−0.23 (0.19)
Musical^b					F(3, 52.72) = 0.94, p = .426	F(3, 58.09) = 0.17, p = .917
Low					−0.16 (0.14)	−0.07 (0.16)
Medium					−0.12 (0.20)	−0.07 (0.20)
High					−0.32 (0.21)	−0.15 (0.23)

Note. Six models were built for activity components: [five individual models] for each component without adjustments for other activity components, and one model with all components corrected for each other. All models are adjusted for antidepressant use. The depressive symptom scores were centered and standardized using the standard deviation and the mean score at baseline. Standard errors are in parentheses. AD = Antidepressant use. p < .05 are in boldface.

^a AD was dummy coded, with ‘no’ as reference category.

^bPhysical, creative, social, and musical activity components were ordinal variables with 4 categories (absent, low, medium, and high), with ‘absent’ as reference category.

^cThe cognitive activity component was an ordinal variable with 3 categories (low, medium, and high), with ‘low’ as reference category.