Neck pain and emotional state in cervical spondylosis: A dual trajectory model analysis

Table 1 Variable assignment

Variables	Assignment
Gender	1 = male, 2 = female
Level of education	1 = elementary school or below, 2 = junior high school, 3 = high school, 4 = college degree, 5 = bachelor's degree or above
Occupation	1 = farmer, 2 = worker, 3 = student, 4 = company clerk, 5 = others
Treatment options	1 = acupuncture treatment, 0 = medication

Table 2 Screening indicators of Neck Pain Questionnaire score trajectory model

Model	BIC (n = 2832)	BIC (n = 472)	AIC	LL	Entropy
Linear
1 set	-12401.78	-12399.09	-12392.86	-12389.86	-
2 sets	-11669.82	-11664.44	-11651.97	-11645.97	0.89
3 sets	-11355.83	-11347.77	-11329.06	-11320.06	0.90
4 sets	-11226.21	-11215.46	-11190.52	-11178.52	0.89
Conic curve
1 set	-12366.22	-12362.63	-12354.32	-12350.32	-
2 sets	-11590.01	-11582.84	-11566.21	-11558.21	0.90
3 sets	-11240.17	-11229.42	-11204.48	-11192.48	0.92
4 sets	-11088.25	-11073.92	-11040.66	-11024.66	0.90
Cubic curve
1 set	-12303.67	-12299.19	-12299.19	-12283.8	-
2 sets	-11435.43	-11426.47	-11405.69	-11395.69	0.91
3 sets	-10999.25	-10985.81	-10954.63	-10939.63	0.93
4 sets	-10790.52	-10772.6	-10731.03	-10711.03	0.92

BIC: Bayesian Information Criterion; AIC: Akaike Information Criterion; LL: Log-likelihood.

Table 3 Results of the Neck Pain Questionnaire score trajectory parameter estimation

Groups	Parameters	Estimated value	Standard error	T score	P value
Low	Intercept	22.170	0.752	29.462	< 0.001
	Linear	-5.624	0.666	-8.447	< 0.001
	Quadratic curve	1.104	0.140	7.861	< 0.001
	Cubic curve	-0.052	0.008	-6.798	< 0.001
Medium	Intercept	45.211	0.681	66.393	< 0.001
	Linear	-10.425	0.578	-18.046	< 0.001
	Quadratic curve	2.138	0.122	17.582	< 0.001
	Cubic curve	-0.106	0.007	-15.839	< 0.001
High	Intercept	69.448	0.919	75.597	< 0.001
	Linear	-14.400	0.816	-17.654	< 0.001
	Quadratic curve	2.986	0.172	17.392	< 0.001
	Cubic curve	-0.148	0.009	-15.676	< 0.001
Group members (%)
Low		33.250	2.309	14.397	< 0.001
Medium		44.527	2.436	18.276	< 0.001
High		22.223	2.002	11.102	< 0.001

Table 4 Distribution of characteristics in each trajectory group of Neck Pain Questionnaire score

Variables	Value	Group (%)			F/χ2	P value
		Low	Medium	High
Age		34.23 ± 15.53	41.02 ± 13.91	43.68 ± 14.88	15.592	< 0.001
Treatment options	0	73 (46.20)	109 (52.66)	69 (64.49)	8.605	0.014
	1	85 (53.80)	98 (47.34)	38 (35.51)
Gender	2	78 (49.37)	108 (52.17)	51 (47.66)	0.642	0.725
	1	80 (50.63)	99 (47.83)	56 (52.34)
Career	1	14 (8.86)	47 (22.71)	35 (32.71)	55.274	< 0.001
	2	14 (8.86)	27 (13.04)	15 (14.02)
	3	63 (39.87)	31 (14.98)	13 (12.15)
	4	25 (15.82)	28 (13.53)	13 (12.15)
	5	42 (26.58)	74 (35.75)	31 (28.97)
Level of education	1	7 (4.43)	14 (6.76)	7 (6.54)	23.361	0.003
	2	30 (18.99)	56 (27.05)	31 (28.97)
	3	21 (13.29)	44 (21.26)	27 (25.23)
	4	8 (5.06)	11 (5.31)	9 (8.41)
	5	92 (58.23)	82 (39.61)	33 (30.84)

Table 5 Factors influencing Neck Pain Questionnaire score trajectory grouping

Medium-level group	Regression coefficients	SE	z score	Wald χ2	P value	OR	OR with 95%CI
Treatment options	-0.414	0.220	-1.879	3.530	0.060	0.661	0.429-1.018
Age	0.033	0.011	3.040	9.242	0.002	1.034	1.012-1.057
Career	-0.028	0.084	-0.329	0.108	0.742	0.973	0.824-1.148
Level of education	-0.019	0.121	-0.156	0.024	0.876	0.981	0.774-1.244
Intercept	-0.613	0.746	-0.821	0.674	0.412	0.542	0.125-2.340
High-level group
Treatment options	-0.961	0.273	-3.525	12.427	< 0.001	0.382	0.224-0.653
Age	0.056	0.014	4.015	16.12	< 0.001	1.058	1.029-1.087
Career	-0.187	0.101	-1.861	3.463	0.063	0.829	0.681-1.010
Degree of education	0.150	0.159	0.943	0.888	0.346	1.161	0.851-1.586
Intercept	-2.084	0.977	-2.132	4.545	0.033	0.124	0.018-0.845

Table 6 Evaluation indicators for fitting the trajectory model of the Short-Form McGill Pain Questionnaire affective item score

Model	BIC (n = 2832)	BIC (n = 472)	AIC	LL	Entropy
Linear
1 set	-6029.91	-6027.22	-6020.98	-6017.98	-
2 sets	-5384.25	-5378.87	-5366.40	-5360.40	0.87
3 sets	-5000.66	-4992.60	-4973.90	-4964.90	0.92
4 sets	-4858.84	-4848.09	-4823.14	-4811.14	0.88
Conic curve
1 set	-5986.74	-5983.16	-5974.85	-5970.85	-
2 sets	-5293.68	-5286.51	-5269.89	-5261.89	0.88
3 sets	-4859.95	-4849.20	-4824.26	-4812.26	0.93
4 sets	-4630.62	-4616.29	-4583.03	-4567.03	0.99
Cubic curve
1 set	-5955.26	-5950.78	-5940.38	-5935.38	-
2 sets	-5225.70	-5216.74	-5195.96	-5185.96	0.89
3 sets	-4750.45	-4737.01	-4705.83	-4690.83	0.93
4 sets	-4521.59	-4503.67	-4462.10	-4442.10	0.98

BIC: Bayesian Information Criterion; AIC: Akaike Information Criterion; LL: Log-likelihood.

Table 7 Parameter estimation results of Short-Form McGill Pain Questionnaire affective item score trajectory

Groups	Parameters	Estimated value	SE	T value	P value
Low	Intercept	1.238	0.112	11.017	< 0.001
	Linear	-1.193	0.104	-11.437	< 0.001
	Quadratic curve	0.209	0.022	9.603	< 0.001
	Cubic curve	-0.010	0.001	-8.132	< 0.001
Medium	Intercept	4.069	0.089	45.869	< 0.001
	Linear	-1.212	0.069	-17.593	< 0.001
	Quadratic curve	0.210	0.014	14.549	< 0.001
	Cubic curve	-0.010	0.001	-12.287	< 0.001
High	Intercept	6.764	0.137	49.355	< 0.001
	Linear	-0.846	0.121	-6.962	< 0.001
	Quadratic curve	0.151	0.026	5.916	< 0.001
	Cubic curve	-0.007	0.001	-5.102	< 0.001
Group members (%)
Low		31.428	2.536	12.408	< 0.001
Medium		52.161	2.628	19.838	< 0.001
High		16.411	1.779	9.224	< 0.001

Table 8 Distribution of characteristics in each trajectory group of the Short-Form McGill Pain Questionnaire affective item score

Variables	Value	Group (%)			F/χ2	P value
		Low	Medium	High
Age		35.53 ± 15.81	40.88 ± 14.98	41.56 ± 13.11	6.884	0.001
Treatment options	0	63 (43.45)	137 (54.80)	51 (66.23)	11.048	0.004
	1	82 (56.55)	113 (45.20)	26 (33.77)
Gender	2	82 (56.55)	120 (48.00)	35 (45.45)	3.518	0.172
	1	63 (43.45)	130 (52.00)	42 (54.55)
Career	1	16 (11.03)	60 (24.00)	20 (25.97)	38.186	< 0.001
	2	15 (10.34)	34 (13.60)	7 (9.09)
	3	55 (37.93)	43 (17.20)	9 (11.69)
	4	24 (16.55)	32 (12.80)	10 (12.99)
	5	35 (24.14)	81 (32.40)	31 (40.26)
Level of education	1	5 (3.45)	20 (8.00)	3 (3.90)	10.747	0.216
	2	33 (22.76)	63 (25.20)	21 (27.27)
	3	25 (17.24)	53 (21.20)	14 (18.18)
	4	6 (4.14)	15 (6.00)	7 (9.09)
	5	76 (52.41)	99 (39.60)	32 (41.56)

Table 9 Factors influencing the grouping of the Short-Form McGill Pain Questionnaire affective item score trajectories

Medium level group	Regression coefficients	SE	z score	Wald χ2	P value	OR	OR with 95%CI
Treatment options	-0.562	0.218	-2.584	6.675	0.010	0.570	0.372-0.873
Gender	0.368	0.219	1.676	2.808	0.094	1.444	0.940-2.220
Age	0.028	0.008	3.700	13.691	< 0.001	1.029	1.013-1.044
Career	0.003	0.076	0.038	0.001	0.970	1.003	0.863-1.165
Intercept	-0.427	0.454	-0.941	0.885	0.347	0.652	0.268-1.588
High level group
Treatment options	-1.091	0.304	-3.591	12.899	0.000	0.336	0.185-0.609
Gender	0.532	0.300	1.775	3.150	0.076	1.703	0.946-3.066
Age	0.036	0.010	3.538	12.516	< 0.001	1.037	1.016-1.057
Career	0.106	0.101	1.048	1.097	0.295	1.112	0.912-1.355
Intercept	-2.139	0.628	-3.406	11.602	0.001	0.118	0.034-0.403

Table 10 Probability (%) of Short-Form McGill Pain Questionnaire trajectory group based on Neck Pain Questionnaire trajectory group condition

NPQ score grouping	Short-form McGill affective item score grouping
	Low	Medium	High
Low	69.1	21.8	9.1
Medium	15.9	69.8	14.3
High	5.4	62.6	32.0

NPQ: Neck Pain Questionnaire.

Table 11 Probability of Neck Pain Questionnaire trajectory group based on Short-Form McGill Pain Questionnaire trajectory group condition (%)

NPQ score grouping	SF-MPQ affective item score grouping
	Low	Medium	High
Low	73.6	14.0	18.6
Medium	22.7	59.8	38.8
High	3.8	26.2	42.6

SF-MPQ: Short-Form McGill Pain Questionnaire; NPQ: Neck Pain Questionnaire.

Table 12 Joint probability of Neck Pain Questionnaire and Short-Form McGill Pain Questionnaire trajectory groups (%)

NPQ score group	SF-MPQ affective item score grouping
	Low	Medium	High
Low	23.1	7.3	3.0
Medium	7.1	31.2	6.4
High	1.2	13.7	7.0

SF-MPQ: Short-Form McGill Pain Questionnaire; NPQ: Neck Pain Questionnaire.