Trajectory and influencing factors of changes in anxiety and depression in elderly patients after lumbar interbody fusion

Abstract

BACKGROUND

Lumbar interbody fusion (LIF) is the primary treatment for lumbar degenerative diseases. Elderly patients are prone to anxiety and depression after undergoing surgery, which affects their postoperative recovery speed and quality of life. Effective prevention of anxiety and depression in elderly patients has become an urgent problem.

AIM

To investigate the trajectory of anxiety and depression levels in elderly patients after LIF, and the influencing factors.

METHODS

Random sampling was used to select 239 elderly patients who underwent LIF from January 2020 to December 2024 in Shenzhen Pingle Orthopedic Hospital. General information and surgery-related indices were recorded, and participants completed measures of psychological status, lumbar spine dysfunction, and quality of life. A latent class growth model was used to analyze the post-LIF trajectory of anxiety and depression levels, and unordered multi-categorical logistic regression was used to analyze the influencing factors.

RESULTS

Three trajectories of change in anxiety level were identified: Increasing anxiety (n = 26, 10.88%), decreasing anxiety (n = 27, 11.30%), and stable anxiety (n = 186, 77.82%). Likewise, three trajectories of change in depression level were identified: Increasing depression (n = 30, 12.55%), decreasing depression (n = 26, 10.88%), and stable depression (n = 183, 76.57%). Regression analysis showed that having no partner, female sex, elevated Oswestry dysfunction index (ODI) scores, and reduced 36-Item Short Form Health Survey scores all contributed to increased anxiety levels, whereas female sex, postoperative opioid use, and elevated ODI scores all contributed to increased depression levels.

CONCLUSION

During clinical observation, combining factors to predict anxiety and depression in post-LIF elderly patients enables timely intervention, quickens recovery, and enhances quality of life.

Key Words: Lumbar interbody fusion; Elderly patients; Anxiety; Depression; Trajectory of change; Influencing factors

Core Tip: A latent class growth model was constructed to analyze trends in postoperative anxiety and depression levels in elderly patients who had undergone lumbar interbody fusion. Three different trends were found: Increasing, decreasing, and stable levels. Factors influencing increased anxiety included female sex, having no partner, high Oswestry dysfunction index (ODI) scores, and low 36-Item Short Form Health Survey scores, while increased depression contributors comprised female sex, postoperative opioid use, and high ODI scores.

INTRODUCTION

With an aging population, the prevalence of spinal disorders in the elderly is increasing, including conditions such as lumbar disc herniation, lumbar spinal stenosis, and lumbar spondylolisthesis[1]. This has led to an increase in the number of elderly patients undergoing lumbar interbody fusion (LIF)[2]. LIF can enhance the structure of the herniated nucleus pulposus in the lumbar intervertebral disc and lessen its pressure on the surrounding nerve roots or spinal cord canal[3,4]. LIF can also alleviate spinal canal stenosis, thereby improving the mechanical environment of the lumbar spine and providing it with enhanced stability[5,6]. However, the recovery time after LIF is long, and patients need to rest in bed or wear lumbar support for protection[7]. Moreover, the procedure somewhat limits the mobility of the lumbar vertebrae, increases stress on neighboring vertebrae, and causes degenerative changes[8]. These physiological outcomes and the pain during recovery can psychologically burden the patient, leading in severe cases to anxiety and depression[9,10]. These consequences typically reduce the patient’s adherence to treatment, leading to a slower postoperative recovery, which prolongs the negative impact on their physical and mental health[11]. To better understand this phenomenon, this study investigated the trajectory of anxiety and depression levels in elderly patients after LIF and the influencing factors. The findings should provide insights into how to reduce postoperative anxiety and depression in this patient population and enhance their recovery speed and quality of life following LIF.

MATERIALS AND METHODS

Patient characteristics

The literature indicates that roughly 12.3% of post-LIF elderly patients are diagnosed with anxiety and/or depression[12]. Applying a tolerance error of 0.05 and alpha value of 0.05 (two-sided test), power analysis using PASS 25 software generated a required sample size of 185 cases. Considering a 20% loss to follow-up rate, a minimum of 232 cases were required. A total of 276 elderly patients who underwent LIF in Shenzhen Pingle Orthopedic Hospital between January 2020 and December 2024 were randomly selected. Three inclusion criteria were applied: (1) Age ≥ 60 years; (2) Successful LIF; and (3) Complete clinical and follow-up information. There were four exclusion criteria: (1) Preoperative severe hepatic or renal insufficiency; (2) Preoperative mental disorder or psychiatric disease; (3) Preoperative malignant tumor; and (4) Inability to complete each scale. The final sample comprised 239 patients. This study was approved by the Ethics Committee of Shenzhen Pingle Orthopedic Hospital (No. PLGSK202400412-3).

Observation indicators

General information: Sex, age, body mass index, education level, place of residence, partner status, underlying medical history, smoking history, drinking history, preoperative diagnosis, preoperative course of the disease, and postoperative use of opioids were recorded.

Surgery-related indicators: The fusion level and number of decompression segments were recorded.

Depressive symptoms: Patients completed the Patient Health Questionnaire-9 (PHQ-9)[13], a tool commonly used for screening, diagnosing, and assessing the severity of depression. Each of its nine items is rated on a 4-point scale (0-3), with a maximum total score of 27 points: 0-4, no depression; 5-9, mild depression; 10-14, moderate depression; 15-19, moderately severe depression; And ≥ 20, severe depression. The scale’s Cronbach’s alpha coefficient was 0.860.

Anxiety symptoms: Patients also completed the Generalized Anxiety Disorder-7 (GAD-7)[14], a widely recognized scale for screening and assessing the severity of anxiety disorders, especially generalized anxiety disorder. Each of the seven items is rated on a 4-point scale, with a maximum total score of 21 points: 0-4, no anxiety; 5-9, mild anxiety; 10-13, moderate anxiety; 14-18, moderately severe anxiety; And ≥ 19, severe anxiety. The scale’s Cronbach’s alpha coefficient was 0.890.

Lumbar spine dysfunction: Patients’ lumbar spine function was measured six months after LIF using the 10-item Oswestry dysfunction index (ODI)[15]. Internationally, the ODI is the most commonly used scale for assessing the degree of functional impairment in patients with low back pain, specifically measuring the impact of pain on their ability to perform activities of daily living. Each item is rated on a 6-point scale, with a maximum total score of 50 points: 0-4 signifies no loss of function; 5-14 mild loss; 15-24 moderate loss; 25-34 severe loss; And ≥ 35 complete loss. The scale’s Cronbach’s alpha coefficient was 0.87.

Quality of life: The 36-Item Short Form Health Survey (SF-36)[16] was used to score patients’ quality of life six months after LIF. The items are organized into eight dimensions, and the maximum total score is 100, with a higher score indicating better quality of life. The scale’s Cronbach’s alpha coefficient was 0.901.

Data collection

For observation indicators “surgery-related indicators” to “quality of life”, data were collected preoperatively and 1 month, 3 months, and 6 months postoperatively.

Statistical analysis

The Mplus8.3 latent class growth model (LCGM) was used to analyze the trajectories of post-LIF anxiety and depression levels. Smaller statistical values indicate a better fit, and the results of trajectory fitting were not considered when the assigned probability was < 10%. SPSS 27.0 was used to process the data, using “mean ± SD” to express normally distributed continuous variables, and “n (%)” to express count information. Intergroup comparisons were performed using an independent samples t-test, χ² test, or Fisher’s precision test. Finally, unordered multi-categorical logistic regression analysis was used to identify the influencing factors of anxiety and depression after LIF.

RESULTS

Potential classes of change trajectories in post-LIF anxiety levels

The specific fitting indicators of the anxiety growth model are shown in Table 1. The Akaike information criterion, Bayesian information criterion (BIC), and adjusted BIC all decreased as the number of model classes increased. With five classes, the smallest group size is 5.5%, which is below the 10% threshold; with four classes, the Lo-Mendell-Rubin-adjusted likelihood ratio test (LMR) is not significant. Considering the fitting indicator values, the LCGM with three classes was considered optimal.

Table 1 Fitting indicators of the latent class growth model for anxiety levels.

No.	AIC	BIC	aBIC	Entropy	LMR	BLRT	Probability
1	3823.352	3844.211	3825.193				1.000
2	3481.818	3513.106	3484.579	0.984	< 0.001	< 0.001	0.121/0.879
3	3334.745	3376.462	3338.426	0.956	< 0.001	< 0.001	0.115/0.119/0.766
4	3251.977	3304.124	3256.578	0.899	0.083	< 0.001	0.114/0.129/0.279/0.477
5	3193.969	3256.546	3199.491	0.924	0.003	< 0.001	0.055/0.065/0.134/0.271/0.476

AIC: Akaike information criterion; BIC: Bayesian information criterion; aBIC: Adjusted Bayesian information criterion; LMR: Lo-Mendell-Rubin-adjusted likelihood ratio test; BLRT: Bootstrap likelihood ratio test.

Trajectories of potential classes of post-LIF anxiety-levels

As shown in Figure 1A, the trajectories of the three potential classes were plotted with GAD-7 scores as vertical coordinates and assessment time points as horizontal coordinates. In class A1, the anxiety score was initially low (I = 2.522, P < 0.001) and showed a gradual increasing trend over time (S = 1.532, P < 0.001); therefore, A1 was named the increasing-anxiety group (n = 26, 10.88%). In class A2, patients’ anxiety score was initially high (I = 8.225, P < 0.001) and tended to decrease over time (S = -1.571, P < 0.001); therefore, A2 was named the decreasing-anxiety group (n = 27, 11.30%). In class 3 (A3), the anxiety score was initially moderate (I = 2.594, P < 0.001), slightly increased, and then decreased again over time (S = 0.139, P = 0.100). The overall score was lower than in groups A1 and A2. Therefore, A3 was named the stable-anxiety group (n = 186, 77.82%). The intercepts of the three groups were 2.522 (P < 0.001), 8.225 (P < 0.001), 2.594 (P < 0.001), and slopes were 1.532 (P < 0.001), -1.571 (P < 0.001), and 0.139 (P = 0.100).

Figure 1 Trajectory of three latent classes after open lumbar interbody fusion. A: Anxiety-level; B: Depression-level. GAD-7: Generalized Anxiety Disorder-7; PHQ-9: Patient Health Questionnaire-9.

Potential classes of change trajectories in post-LIF depression levels

Table 2 shows the fitting indicators for the depression growth model. With five classes, the smallest group size is 5.0%, which is below the 10% threshold; with four classes, the LMR is not significant. Taking into account the fitting indicator values, the linear fit LCGM with three classes was considered the optimal choice.

Table 2 Fitting indicators of the latent class growth model for depression levels.

No.	AIC	BIC	aBIC	Entropy	LMR	BLRT	Probability
1	3864.588	3885.446	3866.428	1.000			1.000
2	3509.856	3541.144	3512.616	0.958	0.034	< 0.001	0.163/0.837
3	3354.116	3395.833	3357.797	0.951	0.007	< 0.001	0.122/0.124/0.754
4	3316.620	3368.767	3321.221	0.843	0.136	< 0.001	0.115/0.120/0.257/0.508
5	3262.497	3325.074	3268.019	0.894	0.051	< 0.001	0.050/0.075/0.113/0.234/0.527

AIC: Akaike information criterion; BIC: Bayesian information criterion; aBIC: Adjusted Bayesian information criterion; LMR: Lo-Mendell-Rubin-adjusted likelihood ratio test; BLRT: Bootstrap likelihood ratio test.

Trajectories of potential classes of post-LIF depression levels

As shown in Figure 1B, the trajectories of the three potential classes were plotted with PHQ-9 scores as vertical coordinates and assessment time points as horizontal coordinates. In class D1, the depression score was initially high (I = 7.657, P < 0.001) and showed a gradual decreasing trend over time (S = -1.243, P < 0.001); therefore, D1 was named the decreasing-depression group (n = 30, 12.55%). In class D2, patients’ anxiety score was initially moderate (I = 2.783, P < 0.001) and tended to increase over time (S = 1.367, P < 0.001); therefore, D2 was named the increasing-depression group (n = 26, 10.88%). In class D3, the depression score was initially lowest (I = 2.288, P < 0.001), slightly increased, and then decreased again over time (S = 0.174, P = 0.027). As the overall score was lower than in groups D1 and D2, D3 was named the stable-depression group (n = 183, 76.57%). The respective intercepts of the three groups were 7.657 (P < 0.001), 2.783 (P < 0.001), 2.288 (P < 0.001), and respective slopes were -1.243 (P < 0.001), 1.367 (P < 0.001), and 0.174 (P = 0.0.027).

Unifactorial analyses of anxiety and depression levels in post-LIF patients

Intergroup comparisons of different anxiety change trajectories (A1, A2, and A3) revealed significant differences among the three groups in terms of sex, partner status, level of fusion, and post- vs pre-surgery change in ODI and SF-36 scores (P < 0.05). The results of intergroup comparisons of different depression change trajectories (D1, D2, and D3) showed significant differences in terms of sex, partner status, postoperative opioid use, and post- vs pre-surgery change in ODI and SF-36 scores (P < 0.05). Table 3 reports the full results of the unifactorial analyses.

Table 3 Univariate analysis of anxiety and depression levels in elderly patients after lumbar interbody fusion, mean ± SD/n (%).

Characteristic	A1 (n = 26)	A2 (n = 27)	A3 (n = 186)	χ2/F value	P value	D1 (n = 30)	D2 (n = 26)	D3 (n = 183)	χ2/F value	P value
Sex				8.320	0.016				6.101	0.047
Male	6 (23.08)	11 (40.74)	97 (52.15)			18 (60.00)	8 (30.77)	101 (55.19)
Female	20 (76.92)	16 (59.26)	89 (47.85)			12 (40.00)	18 (69.23)	82 (44.81)
Age (year)	72.45 ± 5.36	71.85 ± 5.50	73.37 ± 4.04	0.582	0.560	71.47 ± 4.52	72.62 ± 5.78	72.49 ± 5.23	0.653	0.521
BMI (kg/m2)	21.58 ± 1.17	21.59 ± 1.16	21.23 ± 0.88	1.143	0.321	21.22 ± 1.20	21.75 ± 1.11	21.57 ± 1.13	1.633	0.197
Symptom duration				5.479	0.065				4.887	0.087
< 1 year	9 (34.62)	18 (66.67)	92 (49.46)			18 (60.00)	17 (65.38)	84 (45.90)
≥ 1 year	17 (65.38)	9 (33.33)	94 (50.54)			12 (40.00)	9 (34.62)	99 (54.10)
Education level				11.532	0.073				6.003	0.423
Secondary school	14 (53.85)	6 (22.22)	50 (26.88)			9 (30.00)	8 (30.77)	53 (28.96)
Junior high school	6 (23.08)	7 (25.93)	41 (22.04)			7 (23.33)	4 (15.38)	43 (23.50)
High school	3 (11.54)	10 (37.04)	53 (28.49)			10 (33.33)	11 (42.31)	45 (24.59)
College and above	3 (11.54)	4 (14.81)	42 (22.58)			4 (13.33)	3 (11.54)	42 (22.95)
Residence				2.315	0.314				2.702	0.259
City	13 (50.00)	17 (62.96)	88 (47.31)			18 (60.00)	15 (57.69)	85 (46.45)
Countryside	13 (50.00)	10 (37.04)	98 (52.69)			12 (40.00)	11 (42.31)	98 (53.55)
Have a partner				6.250	0.044				7.429	0.024
Yes	9 (34.62)	18 (66.67)	107 (57.53)			19 (63.33)	8 (30.77)	105 (57.38)
No	17 (73.08)	9 (33.33)	79 (42.47)			11 (36.67)	18 (69.23)	78 (42.62)
Smoking				0.250	0.882				5.927	0.052
Yes	13 (50.00)	15 (55.56)	94 (50.54)			11 (36.67)	18 (69.23)	93 (50.82)
No	13 (50.00)	12 (44.44)	92 (49.46)			19 (63.33)	8 (30.77)	90 (49.18)
Alcohol abuse				0.570	0.752				3.252	0.198
Yes	15 (57.69)	13 (48.15)	94 (50.54)			14 (46.67)	9 (34.62)	97 (53.01)
No	11 (42.31)	14 (51.85)	92 (49.46)			16 (53.33)	17 (65.38)	86 (46.99)
History of diabetes				3.348	0.187				1.465	0.481
Yes	17 (65.38)	17 (62.96)	93 (50.00)			19 (63.33)	13 (50.00)	95 (51.91)
No	9 (34.62)	10 (37.04)	93 (50.00)			11 (36.67)	13 (50.00)	88 (48.09)
History of hypertension				1.301	0.522				1.704	0.427
Yes	16 (61.54)	16 (59.26)	96 (51.61)			15 (50.00)	17 (65.38)	96 (52.46)
No	10 (38.46)	11 (40.74)	90 (48.39)			15 (50.00)	9 (34.62)	87 (47.54)
History of coronary heart disease				0.235	0.889				5.089	0.079
Yes	14 (53.85)	15 (55.56)	95 (51.08)			21 (70.00)	11 (42.31)	92 (50.27)
No	12 (46.15)	12 (44.44)	91 (48.92)			9 (30.00)	15 (57.69)	91 (49.73)
Preoperative diagnosis				9.069	0.134				2.877	0.825
Spondylolisthesis	17 (65.38)	10 (37.04)	121 (65.05)			16 (53.33)	15 (57.69)	117 (63.93)
Scoliosis	4 (15.38)	7 (25.93)	24 (12.90)			6 (20.00)	4 (15.38)	25 (13.66)
Recurrent disc herniation	1 (3.85)	2 (7.41)	7 (3.76)			2 (6.67)	1 (3.85)	7 (3.83)
Stenosis	4 (15.38)	8 (29.63)	34 (18.28)			6 (20.00)	6 (23.08)	34 (18.58)
Number of levels decompressed				8.169	0.354				10.085	0.225
1	8 (30.77)	12 (44.44)	63 (33.87)			14 (46.67)	13 (50.00)	56 (30.60)
2	8 (30.77)	6 (22.22)	72 (38.71)			10 (33.33)	4 (15.38)	72 (39.34)
3	3 (11.54)	6 (22.22)	22 (11.83)			3 (10.00)	4 (15.38)	24 (13.11)
4	4 (15.38)	1 (3.71)	19 (10.22)			2 (6.67)	2 (7.69)	20 (10.93)
5	3 (11.54)	2 (7.41)	10 (5.39)			1 (3.33)	3 (11.54)	11 (6.01)
Number of levels fused				11.654	0.020				7.542	0.103
1	8 (30.77)	12 (44.44)	87 (46.77)			10 (33.33)	10 (38.46)	87 (47.54)
2	8 (30.77)	8 (29.63)	74 (39.78)			14 (46.67)	7 (26.92)	69 (37.71)
≥ 3	10 (38.46)	7 (25.93)	25 (13.44)			6 (20.00)	9 (34.62)	27 (14.75)
Chronic opioid use postop				1.002	0.606				6.777	0.034
Yes	14 (53.85)	17 (62.96)	98 (52.69)			11 (36.67)	18 (69.23)	82 (44.81)
No	12 (46.15)	10 (37.04)	88 (47.31)			19 (63.33)	8 (30.77)	101 (55.19)
Post- vs pre-surgery ODI scores				10.926	0.027				10.466	0.032
Elevated	11 (42.31)	7 (25.93)	30 (16.13)			8 (26.67)	11 (42.31)	33 (18.03)
Reduced	6 (23.08)	9 (33.33)	57 (30.65)			11 (36.67)	6 (23.08)	51 (27.87)
No significant change	9 (34.62)	11 (40.74)	99 (53.23)			11 (36.67)	9 (34.62)	99 (54.10)
Post- vs pre-surgery SF-36 scores				10.232	0.033				10.207	0.034
Elevated	7 (26.92)	14 (51.85)	87 (46.77)			12 (40.00)	9 (34.62)	86 (46.99)
Reduced	9 (34.62)	6 (22.22)	24 (12.90)			6 (20.00)	11 (42.31)	27 (14.75)
No significant change	10 (38.46)	7 (25.93)	75 (40.32)			12 (40.00)	6 (23.08)	70 (38.25)

A1: Increasing anxiety; A2: Deceasing anxiety; A3: Stable anxiety; D1: Decreasing depression; D2: Increasing depression; D3: Stable depression; BMI: Body mass index; ODI: Oswestry dysfunction index; SF-36: 36-Item Short Form Health Survey.

Multifactorial analysis of anxiety-level trajectory classes in post-LIF patients

Variables with P < 0.05 in the univariate analyses were included in the unordered multi-categorical logistic regression analyses, which used the latent classes as the dependent variable [with A1 (increased-anxiety) as the reference group]. The regression model likelihood ratio was χ² = 72.001 (P = 0.001). Results showed that patients who had no partner, were female, had elevated ODI scores, or reduced SF-36 scores were more likely to have increased anxiety levels. Table 4 reports the full results.

Table 4 Logistic regression analysis of factors influencing anxiety levels in elderly patients after lumbar interbody fusion.

	β	SE	OR	95%CI	P value
A1 vs A2
Have a partner	1.202	0.595	3.326	1.036-10.682	0.043
A1 vs A3
Sex	-1.186	0.500	0.305	0.115-0.815	0.018
ODI scores	-0.567	0.285	0.567	0.325-0.992	0.047
SF-36 scores	0.596	0.295	1.815	1.012-3.255	0.045

OR: Odds ratio; CI: Confidence interval; A1: Increasing anxiety; A2: Deceasing anxiety; A3: Stable anxiety; ODI: Oswestry dysfunction index; SF-36: 36-Item Short Form Health Survey.

Multifactorial analysis of depression-level trajectory classes in post-LIF patients

The same approach was followed for unordered multi-categorical logistic regression analysis of depression level trajectories, with D2 (increased-depression) as the reference group. The regression model’s likelihood ratio was χ² = 163.655 (P = 0.003). Results showed that patients who were female, postoperative opioid users, or had elevated ODI scores were more likely to have increased depression levels. The full results are reported in Table 5.

Table 5 Logistic regression analysis of factors influencing depression levels in elderly patients after lumbar interbody fusion.

	β	SE	OR	95%CI	P value
D1 vs D2
Sex	-1.291	0.585	0.275	0.087-0.866	0.027
Chronic opioid use postop	-1.413	0.618	0.243	0.073-0.817	0.022
D2 vs D3
Sex	-1.010	0.469	0.364	0.145-0.913	0.031
ODI scores	-0.638	0.265	0.528	0.314-0.888	0.016

OR: Odds ratio; CI: Confidence interval; D1: Decreasing depression; D2: Increasing depression; D3: Stable depression; ODI: Oswestry dysfunction index; SF-36: 36-Item Short Form Health Survey.

DISCUSSION

This study fitted the LCGM to the respective trajectories of anxiety and depression levels of elderly patients after LIF. For anxiety and depression, three groups were identified in which levels were increasing, decreasing, or stable. As the sample included only patients for whom LIF had been successful, the different trends in anxiety and depression levels could be explained by the different cooperative treatments required during the postoperative recovery period and by individual pain perception. Plausibly, new-onset anxiety or depression explains the increasing trends, while preoperative comorbid anxiety or depression explains the decreasing trends. The results of this study are consistent with those of prior research[17].

Unordered multi-categorical logistic regression analysis showed that patients who were female or had no partner, elevated ODI scores, or reduced SF-36 scores were more likely to have increased anxiety levels. Similarly, patients who were female, postoperative opioid users, or had elevated ODI scores were more likely to have increased depression levels. Female elderly patients are generally more emotionally sensitive than their male counterparts, and thus more susceptible to postoperative anxiety and depression[18].

Partner status had a strong influence on the postoperative trajectory of anxiety, with levels increasing over time in patients without a partner. This could be explained by the lack of companionship, depriving patients of a vital outlet for emotional expression and essential psychological support as they cope with the stress of postoperative recovery[19,20]. In an earlier study of middle-aged and older adults, widowed individuals were found to be at higher risk of depression[21].

Preoperative comorbid anxiety or depressive symptoms have been associated with lower surgical satisfaction and slower recovery after LIF, as well as lower health-related quality of life scores[22]. In another study, the postoperative improvement in ODI scores was lower in patients with (vs without) anxiety and depression[23]. Similarly, the present study found that ODI and SF-36 scores six months after surgery were significantly lower in elderly patients with symptoms of anxiety or depression, compared to those without (P < 0.05). The analysis indicates that when elderly patients experience anxiety or depression, their daily routines, regular eating habits, and activities of daily living can be disrupted[24]. Factors such as irregular meal times and inconsistent sleep schedules may cause a decline in postoperative quality of life, which can harm their prognosis. In the opposite direction, for post-LIF patients with poor quality of life scores on the SF-36[25], psychological stress builds up and has fewer outlets for release, thereby raising anxiety levels.

In an earlier study, patient-reported outcome measurement information system scores were lower for opioid users than for non-users[26], suggesting that chronic opioid use is detrimental to patients’ postoperative recovery. This may be explained by a bidirectional association between opioids and both major depressive disorder and anxiety- and stress-related disorders[27]. This study found that postoperative opioid use significantly influenced the development of depression in post-LIF elderly patients, highlighting the need to monitor these patients with follow-up visits and provide timely intervention if they show signs of depression.

In summary, to protect the mental health of elderly patients after LIF, more attention should be given to patients who are female, have no partner, or who, after surgery, use opioids, score higher on the ODI, or score lower on the SF-36. Clinicians should detect these factors promptly by communicating with elderly patients and their families, and implement early interventions to address psychological problems, promote postoperative recovery, and improve quality of life.

However, this study has limitations. Collecting all data from patients of a single hospital may have introduced bias, and the relatively short follow-up period restricts the findings (if the patient’s anxiety and depression levels remain stable within 3 months, but if the follow-up period is extended to 1-3 years, their anxiety and depression levels may show a downward trend). Future research designs should include multiple patient populations and longer follow-up periods to further validate and refine factors influencing the occurrence of anxiety and depression after LIF surgery.

CONCLUSION

This study observed three different trajectories of anxiety and depression levels in post-LIF elderly patients, and identified several factors contributing to increased anxiety or depression over time. Female sex, having no partner, elevated ODI scores, and reduced SF-36 scores are important influencing factors of elevated anxiety levels; somewhat similarly, female sex, postoperative use of opioids, and elevated ODI scores are important influencing factors of elevated depression levels. In clinical practice, these factors can be combined to predict the postoperative occurrence of anxiety and depression in elderly patients, enabling prompt intervention to protect their mental health status, thereby accelerating postoperative recovery and improving patients’ quality of life.