The changing causal foundations of cancer-related symptom clustering during the final month of palliative care: A longitudinal study

Table 2 Indicator Covariance Matrices and Latent Exogenous Variable Correlations

Symptom	Appetite	Tired-ness	Depression	Well-being	Pain	Anxiety	Nausea	Short of Breath	Drowsiness
	One Month Before Death (n = 82)
Appetite	895.3
Tiredness	242.4	399.6
Depress.	296.1	188.5	513.6
Well-being	427.5	218.8	320.6	550.1
Pain	370.8	221.8	231.3	255.7	514.4	0.47	0.41	0.40	0.48
Anxiety	336.0	110.6	388.1	390.7	236.9	577.2	0.50	0.14	0.57
Nausea	172.4	97.6	177.7	144.6	153.3	190.3	297.2	0.16	0.40
Short of Breath	8.5	87.5	43.0	70.3	181.2	66.3	54.6	443.5	0.19
Drowsy	353.4	332.4	301.2	321.4	227.3	275.9	143.0	85.0	505.8
	One Week Before Death (n = 82)
Appetite	1029.2
Tiredness	318.4	623.3
Depress.	250.5	261.9	848.6
Well-being	426.3	262.9	446.2	549.8
Pain	166.4	310.1	276.5	226.1	715.9	0.40	0.37	0.42	0.43
Anxiety	205.2	241.8	661.9	372.4	280.1	811.5	0.65	0.33	0.49
Nausea	150.7	177.6	423.6	300.1	234.4	423.5	620.8	0.39	0.39
Short of Breath	204.4	164.4	347.7	227.4	288.5	243.8	252.2	739.6	0.32
Drowsy	406.6	522.2	359.0	376.6	295.1	345.6	250.5	226.7	765.1

*Indicator variances on the diagonal, indicator covariances below the diagonal, and correlations of the corresponding latent exogenous variables obtained from LISREL's standardized phi matrix in bold above the diagonal. Correlations between any pair of latent indicators can be calculated from the lower triangular entries by dividing the covariance between the indicators by the square root of the product of the two relevant variances $Cov (XY) / \sqrt{Var(X) Var(Y)}$

ISSN: 1471-2288