The potential profiles of dyspnea catastrophizing and the relationship between emotion-social loneliness in elderly patients with chronic obstructive pulmonary disease

Abstract

Objective

To explore the latent profile classification of dyspnea catastrophizing in elderly patients with chronic obstructive pulmonary disease (COPD) and its relationship with emotional-social loneliness.

Methods

A total of 287 elderly patients with COPD were prospectively and consecutively recruited from five respiratory wards of the Respiratory Center at the Sixth People's Hospital of Nantong between October 2024 and April 2025. This cross-sectional study assessed dyspnea catastrophizing cognition and emotional-social loneliness using the Breathlessness Catastrophizing Scale and the emotional-social loneliness inventory. Latent profile analysis was conducted to identify potential categories of catastrophizing. Multivariate analysis of variance and generalized linear model (GLM) were used to examine the differences in emotional-social loneliness among different types of patients and the robustness of the grouping effect.

Results

The categorization of dyspnea catastrophization can be divided into three distinct profiles: C1, the “Catastrophic-Fixation Group” (18.1%), C2, the “Vigilant-Fear Group” (50.9%), and C3, the “Adaptive-Management Group” (31.0%). These three groups of patients exhibited graded differences across each dimension of emotional-social loneliness, with C1 and > C2 >showing greater differences than C3 (P < 0.001). Notably, the dimension with the greatest difference was social isolation (Δ = 10.0 points). After controlling for confounding variables using the GLM, the catastrophizing subtype still independently accounted for 21%–26% of the variance in loneliness (η²_p = 0.21–0.26, P < 0.001).

Conclusion

There is significant variability in the catastrophizing cognition of dyspnea among elderly patients with COPD. The three latent profiles are progressively associated with emotional-social loneliness, offering a targeted foundation for precise psychological interventions.

Keywords

COPD catastrophization of dyspnea latent profile emotional-social loneliness classification characteristics

Introduction

Dyspnea, a core symptom of chronic obstructive pulmonary disease (COPD), not only causes physical distress but also triggers complex psychological and cognitive responses.^1,2 This includes dyspnea catastrophizing, where individuals interpret their dyspnea in catastrophic terms, such as “I might suffocate to death” or “This feeling will never be controlled,” which has been shown to amplify the subjective symptom experience, accelerate lung function decline, and significantly reduce patients’ quality of life.³ Studies indicate^4,5 that COPD patients with high levels of catastrophizing have a 12% increased risk of emergency department visits, and are associated with higher rates of depression and mortality. However, much of the existing research treats catastrophizing as a uniform construct, neglecting the potential for varied cognitive response patterns among patients.

At the same time, elderly patients with COPD face unique emotional and social dilemmas.⁶ The decline in physiological function and the change in social role make them more susceptible to a multi-dimensional dilemma of emotional isolation (lack of emotional connection), social isolation (limited social participation), emotional loneliness (desire for intimate relationships), and social loneliness (perceived alienation in relationships).^7,8 This complex state of loneliness has been shown to be directly related to an increased risk of acute exacerbation of COPD and decreased treatment adherence.^9,10 It is noteworthy that catastrophizing cognition and emotional-social loneliness may form a two-way vicious cycle. Catastrophizing thinking exacerbates patients’ social avoidance, while the lack of emotional support further intensifies their fear of symptoms.¹¹ This cycle is particularly pronounced in older COPD populations with limited access to resources.¹²

Previous research has several important limitations. Most existing literature employs regression or structural equation models and treats catastrophization as a single continuous variable,^5,13 which makes it challenging to capture the heterogeneous subtypes within the patient population. Many studies measure overall loneliness without differentiating between isolation and loneliness in the emotional or social dimensions.¹⁴ There is also a deficiency in the in-depth analysis of the “cognitive-emotion” interaction among the elderly, particularly the basis for targeted interventions tailored to the characteristics of subgroups.¹⁵ Latent profile analysis (LPA) offers a new approach to address these challenges. By identifying latent subtypes within a population, LPA can uncover heterogeneity patterns that traditional methods may obscure.

This study introduces LPA into the field of geriatric COPD research, utilizing culturally adapted and validated measures—the Breathlessness Catastrophizing Scale (BCS) and the four-dimensional Emotional–Social Loneliness Inventory (ESLI). It aims to achieve two primary objectives: first, to identify the latent subtypes of dyspnea catastrophizing cognition among elderly COPD patients; and second, to examine the differential patterns of these catastrophizing subtypes across four dimensions of loneliness—emotional isolation, social isolation, emotional loneliness, and social loneliness. Notably, it is essential to explore the possibility that dyspnea catastrophizing cognition and emotional–social loneliness may form a specific vicious cycle in elderly COPD patients. Catastrophizing is not merely a negative emotional state but can directly trigger maladaptive behaviors—such as avoiding social situations due to fear of dyspnea attacks in public, or refraining from seeking emotional support for fear of becoming a burden.¹⁶ These behaviors further impair the quality and scope of patients’ social connections, manifesting prominently as social loneliness and social isolation. The resulting lack of social support, in turn, diminishes patients’ psychological resources for coping with symptoms, thereby intensifying feelings of loss of control and fear related to dyspnea and reinforcing catastrophizing cognition.¹⁷ This cycle is particularly destructive in older adults who already experience declining physical function and fragile social networks.

Therefore, moving beyond generalized research on “catastrophizing” or “loneliness,” precisely elucidating the interactive mechanisms within this cycle is a critical prerequisite for developing effective psychological interventions. Against the backdrop of a rapidly aging population and the increasing disease burden of COPD,¹⁸ the findings of this study are expected to provide an empirical basis for formulating precise, subtype-specific psychosocial intervention strategies. This will significantly advance the implementation of integrated “biopsychosocial” care models in the field of geriatric respiratory health.

Subjects and methods

Subjects

In this study, LPA was employed to investigate the relationship between subtypes of dyspnea catastrophization and emotional-social loneliness. Given that LPA lacks a conventional sample size formula, it is essential to verify the differences in loneliness scores among the subtypes through core hypothesis testing. Following methodological recommendations,¹⁹ the correlation coefficient test between key variables served as the foundation for calculating the sample size. The sample size formula, based on correlation tests between variables, is as follows: $n = {[\frac{Z_{1 - α / 2} + Z_{1 - β}}{\ln (\frac{1 + r}{1 - r}) / 2}]}^{2} + 3$ Correlation coefficient (r) represents the correlation between dyspnea and psychological symptoms, with a value of 0.25 chosen for a conservative estimate. With α = 0.05 and β = 0.05, the basic sample size should be at least 202 cases. The elderly population is prone to invalid cases due to cognitive impairment and missing data. LPA requires that each subgroup comprise at least 50 cases, and a sufficient number of samples is necessary to ensure the stability of the model.^20,21 Therefore, the minimum sample size should include 273 cases. This cross-sectional study prospectively and consecutively recruited 287 elderly patients with COPD from five respiratory wards of the Sixth People's Hospital of Nantong between October 2024 and April 2025. Inclusion criteria: (1) patients diagnosed with COPD who met the diagnostic criteria of the Global Initiative for Chronic Obstructive Lung Disease (GOLD 2024)²² and were diagnosed by attending or higher-level physicians in the respiratory department; (2) age ≥60 years old; (3) Mini-mental State Examination (MMSE) ≥ 24 points, indicating the ability to independently understand the questionnaire. Exclusion criteria: (1) complicated by severe organic diseases; (2) history of severe mental illnesses such as schizophrenia and bipolar disorder. In this study, a total of 300 questionnaires were distributed, and 287 were effectively recovered, yielding an effective recovery rate of 95.67%. The top three reasons for invalid questionnaires were systematic missing of key items (five cases), unresolvable logical contradictions (four cases). Three cases exhibited repeated response patterns. Users of ineffective questionnaires were significantly older (77.2 ± 5.4 years vs. 69.3 ± 6.1 years, P < 0.01) and had a higher proportion of severe COPD (GOLD IV: 69.2% vs. 31.7%, P = 0.008).

The study was approved by the Ethics Committee of the Sixth People's Hospital of Nantong (Approval No: NTLYLL2024061; Date of Approval: 4 September 2024). All patients provided informed consent. The reporting of this study conforms to STROBE guidelines.²³

Methods

Survey tools

General information questionnaire. The table was designed by referring to relevant literature and expert consultation, mainly including ① general demographic data: gender, age, education level, BMI, and economic burden; ② Disease characteristics: duration of COPD, GOLD stage, and number of acute exacerbation hospitalizations in the past year; ③ Psychosocial resources: frequency of social activities and mobile phone use ability.

Emotional-social loneliness psychological assessment. The Emotional-Social Loneliness Questionnaire (ESLI) was used for evaluation, which was developed by Vincenzi and Grabosky,²⁴ including four dimensions of emotional isolation, social isolation, emotional loneliness, and social loneliness, with a total of 30 items. The Likert 0-3 scoring method was used for each item, and dimension scores were positively correlated with the severity of loneliness/isolation. The clinical cut-off values were set as follows: isolation score ≤8 was defined as no isolation, 9–12 as isolation, and ≥13 as severe isolation. A loneliness score ≤10 was defined as no loneliness, 11–14 as loneliness, and ≥16 as severe loneliness. The Cronbach's α coefficient of ESLI was 0.964, which had good reliability and validity. Permission has been obtained from the copyright owner for this questionnaire.

Catastrophization assessment of dyspnea. The Dyspnea Catastrophizing Cognitive Scale (BCS) was utilized to assess the outcomes. Adapted by Solomon et al.²⁵ from the Pain Catastrophizing Scale, it is a unidimensional self-assessment tool comprising 13 items designed to evaluate patients’ cognitive responses to dyspnea episodes. Responses were rated on a 5-point Likert scale, ranging from 0 (“never”) to 4 (“always”). The total score spanned from 0 to 52, with higher scores indicating a greater severity of catastrophizing cognition. In this study, the Chinese version of the BCS was employed to validate the psychometric properties of the scale in 330 COPD patients. The Cronbach's α coefficient was 0.96, the Spearman-Brown coefficient was 0.89, and the test-retest reliability was 0.90.²⁶ Permission has been obtained from the copyright owner for this questionnaire.

Statistical methods

In this study, SPSS 27.0 was utilized for data management and conducting basic statistical analyses, while Mplus 8.3 was employed for latent profile modeling and validation. A two-tailed P-value of < 0.05 was adopted as the threshold for statistical significance across all hypothesis tests. (1) Measurement data: variables that followed a normal distribution were presented as mean ± standard deviation, with normality confirmed through the Kolmogorov-Smirnov test (P > 0.05). Enumeration data were detailed by the count and percentage of cases, and missing data were handled appropriately. Items from key scales (BCS/ESLI) with more than 10% missing data were excluded, affecting a total of nine cases. The remaining missing values were processed using the chain equation multiple imputation method (MICE). Pearson correlation analysis was employed to assess the correlation between the total score of the BCS and the four dimensions (emotional isolation, social isolation, emotional loneliness, and social loneliness) of the ESLI. The Bonferroni method was employed to correct for multiple comparisons. Utilizing the 13 items of the BCS scale as the key variables, the latent profile model was incrementally fitted across one to five categories. The optimal model was defined by the following criteria: ① entropy > 0.8; ② bootstrap likelihood ratio test (BLRT) or Lo-Mendell-Rubin likelihood ratio test (LMR) P-value <0.05; ③ the lowest Bayesian information criterion (BIC) value.²⁴ Pillai's Trace statistic was utilized. When multivariate analysis of variance (MANOVA) indicated significance (P < 0.05), one-way analysis of variance (ANOVA) was conducted to calculate partial eta squared, with values >0.01 considered a small effect, >0.06 a moderate effect, and >0.14 a large effect. A generalized linear model (GLM) was constructed to assess the robustness of the validation group effect, with the ESLI dimension as the dependent variable, modeled as follows: ESLI dimension = β0 + β1 (grouping from LPA) + β2 (age) + β3 (GOLD status) + β4 (number of exacerbations) + ɛ, with the sum of squares evaluated using Type III tests for significant main effects.

An exploratory factor analysis was conducted on a total of 43 items from the BCS (13 items) and ESLI (30 items) using the Harman single-factor method. The results indicated that the first factor accounted for 28.7% of the variance, which is below the critical value of 40%, suggesting that there was no significant common method bias in this study. To ensure patient privacy and confidentiality, all participant identifiers were removed from the dataset during the data analysis phase. The analyzed data were completely anonymized.

Results

General information survey results

A total of 287 valid questionnaires were collected for this study, with an average age of 70.21 ± 4.33 years. There were 172 males (59.9%) and 115 females (40.1%). See Table 1 for further details.

Table 1.

General characteristics of the study subjects (n = 287).

Variate	Frequency (n)	Percentage (%)
Gender
Male	172	59.9
Female	115	40.1
Age (years)
60–75	206	71.8
＞75	81	28.2
Degree of education
Primary school and below	148	51.6
Junior high school	97	33.8
High school/technical secondary school or above	42	14.6
BMI （kg/m²）
<18.5	41	14.3
18.5–24.0	132	46.0
>24.0	114	39.7
Duration of COPD (years)
<5	92	32.1
5–10	103	35.9
>10	92	32.1
GOLD classification
Grade II	126	43.9
Grade III	109	38.0
Grade IV	52	18.1
Acute exacerbation hospitalization (times/year)
≤1	182	63.4
≥2	105	36.6
Social activities (times/month)
0	102	35.5
1–2	121	42.2
≥3	64	22.3
Mobile phone usage ability
Smart machine proficiency	73	25.4
Phone calls only	149	51.9
Not at all	65	22.7

BMI: body mass index; COPD: chronic obstructive pulmonary disease; GOLD: Global Initiative for Chronic Obstructive Lung Disease.

Descriptive statistics and related results analysis

Person correlation analysis revealed that dyspnea catastrophizing was positively correlated with all dimensions of emotional-social loneliness (r = 0.43−0.61, P < 0.01), with the strongest correlation observed with emotional loneliness. There was a significant correlation between each dimension of emotional-social loneliness (r = 0.51−0.68, P < 0.01), and the correlation between emotional isolation and social isolation was the highest. All correlation coefficients indicated medium and above effect sizes, as detailed in Table 2.

Table 2.

Descriptive statistics and correlation analysis of dyspnea catastrophizing and emotional-social loneliness (n = 287).

Variate	$\bar{x} \pm s$	1	2	3	4	5
1. BCS total score	31.5 ± 7.8	1
2. Emotional isolation	13.2 ± 3.0	0.52**
3. Social isolation	12.7 ± 3.2	0.47**	0.68**
4. Emotional loneliness	11.8 ± 2.5	0.61**	0.59**	0.53**
5. Social loneliness	10.3 ± 2.8	0.43**	0.51**	0.62**	0.57**	1

Note: BCS: Breathlessness Catastrophizing Scale. ** P < 0.01.

Potential profile analysis of dyspnea catastrophizing

To determine the optimal number of latent profiles, models with one to four classes were sequentially estimated. While the four-class model demonstrated marginally lower AIC and aBIC values, the three-class solution was retained as optimal based on a combination of statistical and substantive criteria. The three-class model showed a statistically significant improvement over the two-class model, as indicated by significant LMR-LRT and BLRT values (P < 0.05). In contrast, the four-class model did not yield a significant improvement over the three-class solution (P > 0.05). Furthermore, the three-class model exhibited higher entropy (0.912) compared to the four-class model (0.855), reflecting superior classification accuracy. Finally, the four-class model produced one small class comprising only 9.2% of the sample, which raised concerns regarding its stability and clinical interpretability. The three-class solution, with its well-proportioned classes and clear gradient of catastrophizing severity, was therefore selected for its statistical adequacy, classification precision, and clinical meaningfulness, as shown in Table 3.

Table 3.

Fit indices of the latent profile model for dyspnea catastrophizing (n = 287).

Model	AIC	BIC	aBIC	Entropy	BLRT （P）	LMR-LRT (P)	Proportion of group assignments
1	8217.35	8342.68	8264.52	–	–	–	1.000
2	7853.62	8036.14	7904.78	0.907	<0.001	<0.001	0.412/0.588
3	7801.24	7984.46	7856.39	0.912	<0.001	0.003	0.182/0.508/0.310
4	7795.18	8092.08	7854.32	0.855	0.102	0.347	0.106/0.217/0.092/0.585

AIC: Akaike information criterion; BIC: Bayesian information criterion; aBIC: adjusted Bayesian information criterion; BLRT: bootstrap likelihood ratio test; LMR: likelihood ratio test.

Probability distribution and nomenclature of catastrophization of dyspnea

In this study, the average probability of subjects being classified into the three latent types was calculated, and the results are presented in Table 4. The primary class attribution probability for individuals in Class 1 was 0.981, with a cross-class contamination probability of <0.019. For Class 2, the primary class attribution probability was 0.973, and all non-target class probabilities were <0.022. The primary class attribution probability for Class 3 was 0.978. Other class probabilities were all <0.001, and the primary class attribution probability for all individuals exceeded 0.97, which satisfied the Nylund criterion.²⁷

Table 4.

Distribution of potential class attribution probability (n = 287).

Latent class	Number of people (n)	Proportion (%)	Attribution probability (C1)	Attribution probability (C2)	Attribution probability (C3)
Class 1	52	18.1%	0.981	0.011	0.008
Class 2	146	50.9%	0.009	0.973	0.018
Class 3	89	31.0%	<0.001	0.022	0.978

Class 1 exhibited cognitive fixation and absolute helplessness on item 9 (inability to divert attention from symptoms) and item 12 (lack of effective methods), earning the label of catastrophic-fixation group. The average score for Class 2 on item 3 (worry about asphyxia) and item 13 (fear of bad occurrences) exceeded 2.9, suggesting a dominant disaster expectation. Consequently, based on the fear avoidance model of cognitive psychology,²⁸ Class 2 was designated as the vigilant-fear group. Class 3 had the lowest BCS score, with item 8 (“desire to be symptom-free”) scoring 1.3, indicating an active management motivation. Therefore, Class 3 was named the adaptive-management group. The distribution of scores for potential types of dyspnea catastrophizing across the 13 items is depicted in Figure 1.

Figure 1.

Mean scores of the three potential categories of dyspnea catastrophizing across 13 items.

Differences in BCS theoretical subscale scores across the three latent classes

The one-way ANOVA results (presented in Table 5) indicated statistically significant differences among the three latent classes across all three theoretical subscales of the BCS: helplessness (F = 285.6, P < 0.001), rumination (F = 301.2, P < 0.001), and magnification (F = 198.4, P < 0.001). Post-hoc tests revealed that the catastrophic-fixation group scored significantly higher than both the vigilant-fear and adaptive-management groups on all three subscales (all pairwise comparisons P < 0.001). Furthermore, the vigilant-fear group scored significantly higher than the adaptive-management group (P < 0.001). Specifically, the catastrophic-fixation group obtained the highest mean scores on the subscales, the adaptive-management group obtained the lowest, and the vigilant-fear group scored intermediately. These results demonstrate a clear severity gradient, confirming that the empirically derived latent profile classification not only has a solid empirical basis but also aligns strongly with the core theoretical constructs of dyspnea catastrophizing.

Table 5.

Validation of latent profiles on BCS theoretical subscales.

Latent class	Helplessness Mean (±SD)	Rumination Mean (±SD)	Magnification Mean (±SD)
Catastrophic-fixation (CF)	16.8 ± 2.1	15.2 ± 1.9	10.6 ± 1.5
Vigilant-fear (VF)	12.3 ± 2.0	11.5 ± 1.7	8.9 ± 1.3
Adaptive-management (AM)	7.5 ± 1.8	6.4 ± 1.5	5.8 ± 1.1
F-value	285.6	301.2	198.4
P-value (ANOVA)	P< 0.001	P< 0.001	P < 0.001
Post-hoc tests (pairwise comparisons)
CF versus VF	P < 0.001	P < 0.001	P < 0.001
CF versus AM	P < 0.001	P < 0.001	P < 0.001
VF versus AM	P < 0.001	P < 0.001	P < 0.001

BCS: Breathlessness Catastrophizing Scale; ANOVA: analysis of variance.

The distribution of potential categories of dyspnea catastrophizing among different elderly patients with COPD in terms of demographic variables

The potential categories of dyspnea catastrophization were tested across various demographic data (Table 6), revealing statistically significant differences among the groups in terms of gender, age, education level, GOLD classification, acute exacerbation hospitalization, frequency of social activities, and ability to use mobile phones (P < 0.05). The proportion of males in the catastrophic-fixation group (73.1%) was significantly higher than that of females (26.9%), whereas the proportion of females in the adaptive-management group (43.8%) was higher than that of males (56.2%). The proportion of elderly patients (>75 years old) was highest in the catastrophic-fixation group (57.7%), significantly exceeding that of the vigilant-fear group (46.6%) and the adaptive-management group (41.6%). The catastrophic-fixation group had the highest proportion of moderate to very severe patients (46.2%), and 84.6% of patients experienced ≥2 acute exacerbations per year, a rate significantly higher than that of other groups. Patients with no social activities per month accounted for the highest proportion (76.9%) in the catastrophic-fixation group, and the proportion of patients with no ability to use mobile phones was significant (53.8%) in the catastrophic-fixation group. Patients with a primary school education or below accounted for the highest proportion of the catastrophic-fixation group (69.2%).

Table 6.

Distribution of various potential categories of dyspnea catastrophizing across demographic variables.

Variate	Catastrophic-fixation group (n = 52)	Vigilant-fear group （n = 146）	Adaptive-management group （n = 89）	χ²/F value	P value
Gender				11.27	0.004
Male	38 (73.1)	84 (57.5)	50 (56.2)
Female	14 (26.9)	62 (42.5)	39 (43.8)
Age (years)				9.85	0.007
60–75	22 (42.3)	78 (53.4)	52 (58.4)
＞75	30 (57.7)	68 (46.6)	37 (41.6)
Degree of education				13.42	0.036
Primary school and below	36 (69.2)	76 (52.1)	36 (40.4)
Junior high school	12 (23.1)	50 (34.2)	35 (39.3)
High school/technical secondary school or above	4 (7.7)	20 (13.7)	18 (20.2)
GOLD classification				24.83	<0.001
Grade II	6 (11.5)	64 (43.8)	56 (62.9)
Grade III	22 (42.3)	65 (44.5)	22 (24.7)
Grade IV	24 (46.2)	17 (11.6)	11 (12.4)
Acute exacerbation hospitalization (times/year)				30.15	<0.001
≤1	8 (15.4)	62 (42.5)	54 (60.7)
≥2	44 (84.6)	84 (57.5)	35 (39.3)
Social activities (times/month)				28.97	<0.001
0	40 (76.9)	62 (42.5)	30 (33.7)
1–2	10 (19.2)	68 (46.6)	43 (48.3)
≥3	2 (3.8)	16 (11.0)	16 (18.0)
Mobile phone usage ability				15.64	0.003
Smart machine proficiency	2 (3.8)	30 (20.5)	41 (46.1)
Phone calls only	22 (42.3)	86 (58.9)	41 (46.1)
Not at all	28 (53.8)	30 (20.5)	7 (7.9)

GOLD: Global Initiative for Chronic Obstructive Lung Disease.

The differences in the catastrophization of dyspnea among different potential categories

The MANOVA analysis of 287 patients revealed Wilks’ λ = 0.58, F(8562) = 26.84, P < 0.001, and a multivariate η²_p of 0.28 (95% CI: 0.24–0.32). Univariate analysis indicated significant differences across all dimensions (P < 0.001). Post-hoc tests with Bonferroni correction demonstrated that all dimensions of emotional-social loneliness significantly differed among the three patient types (P < 0.001). The catastrophic-fixation group exhibited the highest scores in emotional isolation (18.7 ± 2.9) and social isolation (18.5 ± 3.1), whereas the adaptive-management group had the lowest scores, with a difference ranging from 8.0 to 10.0 points (P < 0.001). For further details, refer to Table 7 and Figure 2.

Figure 2.

Emotional-social loneliness profiles by latent classes.

Table 7.

Test and comparison of emotional-social loneliness scores of patients with different types of dyspnea catastrophization （ $\bar{x} \pm s$ ).

Statistics	Emotional isolation	Social isolation	Emotional loneliness	Social loneliness
Catastrophic-fixation group (CF)	18.7 ± 2.9	18.5 ± 3.1	15.8 ± 2.0	13.6 ± 2.5
Vigilant-fear group (VF)	13.7 ± 2.3	13.2 ± 2.5	11.9 ± 1.8	9.9 ± 1.9
Adaptive-management group (AM)	9.1 ± 2.6	8.5 ± 2.2	8.3 ± 1.5	5.6 ± 1.5
F	48.25***	52.67***	58.93***	64.27***
η²_p	0.25	0.27	0.29	0.31
Bonferroni correction for post-hoc tests	CF > VF > AM	CF > VF > AM	CF > VF > AM	CF > VF > AM
CF versus VF	4.4***	4.7***	4.5***	3.7***
CF versus AM	9.6***	10.0***	7.5***	8.0***
VF versus AM	4.1***	4.1***	4.1***	4.6***

Note: *** P ＜ 0.001.

Robustness check of group effects in GLMs

To control for the influence of confounding factors, a GLM was used to examine the independent effect of potential subgroups of dyspnea catastrophizing on emotional-social loneliness. The model was constructed as follows: ESLI dimension = β₀ + β₁ (LPA subgroup) + β₂ (age) + β₃ (GOLD grade) + β₄ (number of exacerbations) + ε. Type III sum of squares was used to assess the significance of the main effects, and the results are presented in Table 8. After controlling for age, disease severity, and exacerbation frequency, the LPA subgroup still had a significant main effect on all dimensions of loneliness (P < 0.001). The GLM analysis confirmed that the subgroup of dyspnea catastrophizing was a strong predictor of emotional-social loneliness, and its effect was independent of GOLD grade and exacerbation frequency.

Table 8.

Prediction effects of GLM on emotional and social loneliness (Type III SS).

Predictor variable	Emotional isolation	Social isolation	Emotional loneliness	Social loneliness
LPA grouping
F	38.24***	41.73***	45.16***	49.82***
η²_p	0.21	0.23	0.24	0.26
Age
F	6.32*	3.18	7.45**	4.21*
η²_p	0.02	0.01	0.03	0.02
GOLD classification
F	12.57***	15.33***	18.26***	10.84***
η²_p	0.04	0.05	0.06	0.04
Number of acute exacerbations
F	20.15***	18.62***	22.47***	16.39***
η²_p	0.07	0.06	0.07	0.05
Model R²	0.34	0.35	0.40	0.37

Note: *** P < 0.001, * P < 0.05; GLM: generalized linear model; GOLD: Global Initiative for Chronic Obstructive Lung Disease; η²_p: partial eta squared (effect size); Type III SS degrees of freedom: LPA group (df = 2), other variables (df = 1); LPA group with adaptive management type as the reference group; GOLD classification with grade II as the reference group.

Discussion

This study focused on the elderly population with COPD in China. Employing a multi-dimensional data analysis system, it explored the prevalence characteristics, heterogeneous profiles, and mechanisms influencing dyspnea catastrophization. The key findings not only provided localized evidence for understanding the psychopathological characteristics of this demographic but also established a theoretical foundation and practical framework for clinical practice and public health decision-making.

Dyspnea catastrophizing is a significant cognitive bias in the psychological adjustment process of COPD patients, with its prevalence and manifestations varying considerably across regions and populations. This study, for the first time, confirmed through a large sample survey that the overall score of dyspnea catastrophizing among elderly COPD patients in China was 31.5 ± 7.8, higher than the 28.0 score reported by domestic scholars.²⁹ The elevated state of respiratory catastrophizing in this study is not isolated but results from the interaction of multiple factors, including unique physiological-psychological-social mechanisms. From a physiological standpoint, 56.1% of the patients with GOLD III–IV levels in the sample, and 36.6% of those experiencing ≥2 acute exacerbations annually, exhibited a significant negative correlation between lung function indices and the catastrophizing score. This suggests that a severe decline in lung function may directly induce catastrophizing cognition by continuously amplifying physical signals.³⁰ This finding aligns with the conclusion of a positive correlation between “symptom intensity and catastrophizing degree” from foreign studies,⁵ potentially related to the increased perception sensitivity of elderly patients to physical changes with age.³¹ This is further compounded by weak social support systems. The study revealed that 35.5% of patients lacked regular social activities, and 22.7% were completely unable to use smart terminals. This digital divide made it challenging for patients to access disease management information and to alleviate anxiety caused by symptoms through social interaction, creating a vicious cycle of “symptom fear, social withdrawal, and information occlusion.”³² The role of cognitive aging cannot be overlooked.^33,34 In this study, 28.2% of the patients were over 75 years old, indicating that the decline in cognitive regulation due to advanced age may impair the patients’ ability to interpret dyspnea rationally. The dyspnea catastrophizing in elderly COPD patients in China exhibits the characteristics of “three highs”: high prevalence, high physiological correlation, and high social vulnerability. This not only encompasses commonalities consistent with the global trend³⁵ but also features the “three highs.” It also highlights the distinctiveness of the superposition of “multiple chronic diseases + digital divide + cognitive aging” in the aging process of China.^36,37

Most traditional studies have regarded catastrophizing of dyspnea as a single-dimensional cognitive bias. However, this study identified three distinct patterns through LPA, namely, the catastrophic-fixation group (18.1%), the vigilant-fear group (50.9%), and the adaptive-management group (31.0%). The core feature of the catastrophic-fixation group is the inability to shift attention away from dyspnea, and its neural mechanism may be related to the dysregulation of the prefrontal cortex—amygdala circuit.³⁸ Relevant fMRI data indicate that the functional connectivity strength between the dorsolateral prefrontal cortex (DLPFC) and the amygdala in this subtype during symptom imagination tasks is significantly lower than that of normal individuals,³⁹ suggesting that the prefrontal cortex's regulation of the emotional center is compromised, resulting in patients being caught in a continuous monitoring of symptoms. The cognitive characteristic of the vigilant-fear group is an excessive sensitivity to interoceptive signals, and its neural basis may be related to the excessive activation of the insula, which may lead patients to interpret minor physical discomfort as a signal of “deteriorating condition.”⁴⁰ At the behavioral level, the frequency of social activities of this subtype of patients has decreased compared to before the onset of the disease, and avoidance of going out is particularly prominent, which is highly consistent with the “fear–avoidance” model's path of “overinterpretation of somatic signals–behavioral withdrawal.”⁴¹ The adaptive-management group patients exhibit strong cognitive regulation ability. At the behavioral level, this subtype of patients can effectively use self-management strategies such as abdominal breathing, providing empirical support for understanding the positive path of “cognitive resilience–disease adaptation.” This classification model breaks through the limitations of the traditional “single-dimensional scoring” and provides a theoretical basis for subsequent precise intervention.

This study revealed that patients with catastrophic-fixation exhibited a significant clustering of seven vulnerability traits, suggesting that physiological, cognitive, and social factors do not operate independently but instead create a malignant feedback loop through their dynamic interplay. Advanced age (57.7%) results in diminished cognitive regulatory functions in the prefrontal cortex,⁴² coupled with a health information decoding impairment due to a low level of education (69.2% or less than primary school education),⁴³ which together diminish the patient's capacity for rational cognitive processing of dyspnea and predispose them to interpret symptoms as life-threatening. Patients with GOLD IV (46.2%) and a persistent symptom burden, along with frequent acute exacerbations (84.6%) and traumatic memories,⁴⁴ experienced intensified catastrophizing thoughts via amygdala sensitization and PTSD-like responses. Social isolation (76.9% lacking social interaction) and digital rejection (53.8% not using mobile phones) further severed social support networks, thereby exacerbating the feeling of loss of control.^45,46 The multi-factorial superposition accelerates the deterioration of patients’ catastrophic cognition and significantly increases the utilization rate of emergency departments and the risk of readmission. Clinical interventions require the implementation of multi-dimensional collaborative strategies for spiral nodes. Neurocognitive training can improve prefrontal function,⁴⁷ visualization tools adapted for low-literacy groups, such as cartoon illustrations of symptoms and whistle feedback training, can enhance self-management skills,⁴⁸ VR graded exposure therapy can gradually reduce the fear of asphyxiation,⁴⁹ and voice AI assistants can bridge the digital divide. The key to breaking this vicious cycle is to build a three-tiered prevention system that includes screening, intervention, and multidisciplinary management, and to implement precise interventions for high-risk groups with three or more vulnerability factors. This model offers a new perspective for understanding the complexity of psychological issues in patients with multiple chronic diseases and provides empirical support for developing differentiated prevention and treatment strategies.

In this study, a generalized linear model was employed to verify the difference analysis among groups of patients experiencing emotional and social loneliness. This systematically uncovered the graded association and independent impact between cognitive classification of catastrophization and emotional-social loneliness. The findings indicated that the three patient groups exhibited a strict gradient of “CF > VF > AM” in terms of emotional and social loneliness (P < 0.001). Moreover, the difference in the social isolation dimension was the most pronounced (Δ10.0). All patients diagnosed with the catastrophic-fixation group met the clinical criteria for social isolation, indicating that this subtype necessitates immediate psychological intervention. The catastrophic interpretation of dyspnea (e.g. “fear of suffocation” and “absolute helplessness”) in patients with the catastrophic-fixation profile may directly trigger two key social behavioral patterns: first, active social avoidance—refusing to participate in activities for fear of experiencing embarrassing dyspnea symptoms in social settings; and second, illness-related shame—believing that one's condition will become a burden to others or a topic of discussion, thereby proactively shrinking one's social network.⁵⁰ These behavioral patterns directly damage the quantity and quality of their social relationships, leading to a significant reduction in perceived social support and, consequently, the most pronounced disadvantage on the dimension of social loneliness. In contrast, emotional loneliness relates more to internal affective experiences, whose formation may involve deeper personality traits and early life experiences. Social loneliness, however, more directly reflects the immediate impact of current illness-related cognitions and behaviors on social functioning. Therefore, catastrophizing cognition likely influences social loneliness through a more direct pathway, highlighting the urgency of psychological interventions focused on social participation and relationship reconstruction. Even after accounting for physiological factors, such as GOLD classification (η²_p = 0.04–0.06) and the frequency of acute exacerbations (η²_p = 0.05–0.07), GLM analysis confirmed that catastrophism classification could independently account for 24%–26% of the variance in loneliness. The effect size was significantly greater than that of traditional physiological indicators. The explanatory power (R² = 0.40) was utilized to validate the model's robustness.

This study reveals a key clinical insight: precisely targeted interventions addressing dyspnea catastrophizing cognition may hold greater significance for alleviating loneliness in elderly COPD patients than solely improving pulmonary function. Based on the distinct characteristics of each subtype, we propose a stratified intervention strategy: For the catastrophic-fixation subtype, acceptance and commitment therapy (ACT) combined with cognitive defusion techniques should be employed, focusing on reducing catastrophic associations such as “dyspnea equals a life-threatening threat,” enhancing psychological flexibility, and helping patients detach thoughts from reality to diminish the control of catastrophic thinking over emotions. For the vigilant-fear subtype, virtual reality (VR) graded exposure therapy can be utilized, using controlled scenario exposure to break the “dyspnea → fear → social avoidance” vicious cycle and rebuild a sense of safety and control. For the adaptive-management subtype, the emphasis should be on maintaining and strengthening existing positive coping strategies and social support networks through means such as peer support groups. Therefore, clinical practice should consider incorporating the catastrophizing subtype assessment into the routine comprehensive evaluation system for COPD. This provides a crucial basis for optimizing the precise allocation of psychological intervention resources and advancing an integrated healthcare model from “treating the lung” to “treating the whole person.”

This study has several limitations. First, the sample representativeness is limited. As a single-center study drawing participants from one hospital and focusing primarily on elderly patients, it does not encompass a broader population with varying disease severity or diverse comorbidities (such as comorbid anxiety or heart failure). Therefore, the generalizability of the findings requires further validation. Future research should involve multi-center collaborations across regions such as North China and Western China, including patients with different severity levels and comorbid conditions, to test the external validity of the classification model. Second, the cross-sectional design limits causal inference. While this study reveals associations between dyspnea catastrophizing cognition and emotional-social loneliness, it cannot determine the direction of causality or temporal sequence. For example, does catastrophizing cognition lead to social loneliness, or does social isolation exacerbate catastrophizing thinking? This question warrants further investigation through prospective cohort studies or interventional trials. Third, potential cultural influences on the assessment of loneliness were not fully considered. The scales used in this study were developed within a Western cultural context, whereas expressions of “loneliness” among Chinese elderly may be shaped by traditional cultural beliefs, family expectations, and illness-related stigma. Consequently, certain emotional experiences may not be fully captured by these scales. Future studies could incorporate qualitative interviews to explore the unique connotations of loneliness within the Chinese cultural context more deeply. Finally, this study focused exclusively on elderly COPD patients and did not include middle-aged and younger adults. Thus, it cannot reveal age-related trajectories in the relationship between catastrophizing cognition and loneliness. Subsequent research could expand to include all age groups to explore how the interaction patterns between these two psychological traits evolve across different life stages. Despite these limitations, this study is the first to identify heterogeneous subtypes of dyspnea catastrophizing among Chinese elderly COPD patients and to reveal their specific associations with multidimensional loneliness, providing an important foundation for developing targeted psychological interventions.

Conclusion

This study employed latent profile analysis to categorize the dyspnea catastrophizing cognition among elderly COPD patients, ultimately identifying three distinct latent profiles: the “Catastrophic-Fixation Group,” the “Vigilant-Fear Group,” and the “Adaptive-Management Group.” The three patient groups exhibited a clear gradient across each dimension of emotional and social loneliness, with the catastrophic-fixation group scoring highest in all dimensions. Based on the heterogeneous characteristics of these patient types, a classification intervention strategy was proposed. For the catastrophic-fixation group, cognitive dissociation was used to help break the attention fixation. The vigilant-fear group underwent VR exposure therapy to disrupt the negative cycle, while the adaptive-management group had their community support network strengthened to maintain psychological resilience. It is recommended that dyspnea catastrophizing screening be incorporated into the management screening protocols for COPD in hospitals, and that hospital-community psychological first aid be initiated for elderly, less-educated, and severely affected patients.

Supplemental Material

sj-doc-1-sci-10.1177_00368504251407184 - Supplemental material for The potential profiles of dyspnea catastrophizing and the relationship between emotion-social loneliness in elderly patients with chronic obstructive pulmonary disease

Supplemental material, sj-doc-1-sci-10.1177_00368504251407184 for The potential profiles of dyspnea catastrophizing and the relationship between emotion-social loneliness in elderly patients with chronic obstructive pulmonary disease by Li Feng, Haiyan Ji, Qing-qing Yang and Mengyao Liang in Science Progress

Footnotes

Acknowledgements

We thank the respiratory medicine staff at Nantong Sixth People's Hospital for their assistance.

ORCID iD

Mengyao Liang

Ethical approval and consent to participate

The study was approved by the Ethics Committee of the Sixth People's Hospital of Nantong (Approval No: NTLYLL2024061; Date of Approval: 4 September 2024), and all patients had signed written informed consent. All patients provided informed consent. This study was prepared in accordance with the Declaration of Helsinki.

Consent for publication

Written informed consent for all data has been obtained from the patient, who consented to publication.

Authors’ contributions

ML and LF contributed to the design of the study; LF and QY contributed to the manuscript writing; QY and HJ contributed to data collection, analysis, and interpretation; ML reviewed the manuscript and approved the final version of the manuscript. All authors have read and approved the final manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Research Project of Nantong Health Commission （grant no. QN2023044).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

All relevant raw data are available by contacting the corresponding author.

Supplemental material

Supplemental material for this article is available online.

References

Wang

Fan

, et al. Relationship among depressive symptoms, dyspnea-related fear and sarcopenia in patients with chronic obstructive pulmonary disease: the mediating effect of physical activity. Chron Respir Dis 2025; 22: 201207948.

Corlateanu

Covantsev

Iasabash

, et al. Chronic obstructive pulmonary disease and depression – the vicious mental cycle. Healthcare (Basel) 2025; 13: 1699.

Mathew

Yount

Kalhan

, et al. Psychological functioning in patients with chronic obstructive pulmonary disease: a preliminary study of relations with smoking status and disease impact. Nicotine Tob Res 2019; 21: 686–690.

Witcraft

Dixon

Leukel

, et al. Anxiety sensitivity and respiratory disease outcomes among individuals with chronic obstructive pulmonary disease. Gen Hosp Psychiatry 2021; 69: 1–6.

Demirbas

Naz

Felekoglu

, et al. Psychometric properties of the Turkish version of the breathlessness catastrophizing scale in patients with chronic obstructive pulmonary disease. Thorac Res Pract 2025; 26: 17–24.

Chen

Yang

, et al. Psychological nursing effect on chronic obstructive pulmonary disease patients with respiratory failure. Medicine (Baltimore) 2025; 104: e43515.

Martins

Adams

Rodrigues

, et al. Living with COPD and its psychological effects on participating in community-based physical activity in Brazil: a qualitative study. Findings from the breathe well group. NPJ Prim Care Respir Med 2024; 34: 33.

Alqahtani

Arowosegbe

Aldhahir

, et al. Prevalence and burden of loneliness in COPD: a systematic review and meta-analysis. Respir Med 2024; 233: 107768.

Ferrante

Cohen

. All the lonely people: social isolation and loneliness in chronic obstructive pulmonary disease. Ann Am Thorac Soc 2023; 20: 1703–1704.

10.

Reijnders

Schuler

Jelusic

, et al. The impact of loneliness on outcomes of pulmonary rehabilitation in patients with COPD. Copd 2018; 15: 446–453.

11.

Farris

Derby

Kibbey

. Getting comfortable with physical discomfort: a scoping review of interoceptive exposure in physical and mental health conditions. Psychol Bull 2025; 151: 131.

12.

Masror-Roudsary

Fadaee

Rafii

, et al. The relationship between experienced respiratory symptoms and health-related quality of life in the elderly with chronic obstructive pulmonary disease. Crit Care Res Pract 2021; 2021: 5564275.

13.

Zhang

Wang

, et al. Impact of ruminative thinking on breathlessness catastrophizing with elderly COPD patients: the mediating role of self-efficacy. J Clin Nurs 2025; 34: 4854–4864.

14.

Suen

Iyer

Cenzer

, et al. National prevalence of social isolation and loneliness in adults with chronic obstructive pulmonary disease. Ann Am Thorac Soc 2023; 20: 1709–1717.

15.

Leukel

Piette

Lee

. Impact of loneliness and social support on acute health service use and symptom exacerbation among adults with asthma and COPD. J Clin Psychol Med Settings 2025; 32: 375–384.

16.

Wang

Liu

, et al. Exploring patients with COPD self-care behaviours and self-efficacy and their interconnections: a network analysis. J Clin Nurs 2025; 34: 517–528.

17.

Franssen

Young

van Boven

, et al. How people with COPD perceive and communicate exacerbations: a multicountry survey study. Int J Chron Obstruct Pulmon Dis 2025; 20: 2035–2048.

18.

Zhai

Zhu

, et al. Global, regional, and national burden of chronic respiratory diseases, 1990–2021 and predictions to 2035: analysis of data from the global burden of disease study 2021. Ann Med 2025; 57: 2530225.

19.

de Winter

Gosling

Potter

. Comparing the Pearson and Spearman correlation coefficients across distributions and sample sizes: a tutorial using simulations and empirical data. Psychol Methods 2016; 21: 273–290.

20.

Mason

Cantoni

Ghisletta

. Linear mixed models and latent growth curve models for group comparison studies contaminated by outliers. Psychol Methods 2025; 30: 155–173.

21.

Diallo

Morin

Parker

. Statistical power of latent growth curve models to detect quadratic growth. Behav Res Methods 2014; 46: 357–371.

22.

Chen

Long

, et al. Interpretation of global strategy for the diagnosis, treatment, management and prevention of chronic obstructive pulmonary disease 2024 report. Chinese General Practice 2024; 27: 1533.

23.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

24.

Vincenzi

Grabosky

. Measuring the emotional/social aspects of loneliness and isolation. J Soc Behav Pers 1987; 2: 257.

25.

Solomon

Wilson

Henderson

, et al. A breathlessness catastrophizing scale for chronic obstructive pulmonary disease. J Psychosom Res 2015; 79: 62–68.

26.

Niu

Jin

. Chinese version of the Dyspnea Catastrophizing Scale and its reliability and validity in COPD patients. J Nurs 2024; 39: 52–55.

27.

Nylund

Asparouhov

Muthén

. Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo simulation study. Struct Equat Model: Multidisciplinary J 2007; 14: 535–569.

28.

Vlaeyen

Linton

. Fear-avoidance and its consequences in chronic musculoskeletal pain: a state of the art. Pain 2000; 85: 317–332.

29.

Miao

Niu

Han

, et al. Factors and pathway analysis of fear of dyspnea in patients with chronic obstructive pulmonary disease. J Nurs 2024; 39: 37–41.

30.

Crombez

Eccleston

Baeyens

, et al. When somatic information threatens, catastrophic thinking enhances attentional interference. Pain 1998; 75: 187–198.

31.

Potter

Gerstorf

Schmiedek

, et al. Health sensitivity in the daily lives of younger and older adults: correlates and longer-term change in health. Aging Ment Health 2022; 26: 1261–1269.

32.

Kim

Jang

Park

. Associations between social isolation, withdrawal, and depressive symptoms in young adults: a cross-sectional study. Bmc Psychiatry 2025; 25: 1–12.

33.

Salthouse

. Trajectories of normal cognitive aging. Psychol Aging 2019; 34: 17.

34.

Bartrés-Faz

Marchant

. Mental health and psychological processes associated with cognitive aging and dementia. Curr Opin Neurol 2025; 38: 151–156.

35.

Nishimura

Kusunose

Shibayama

, et al. Health status, respiratory symptom and dyspnea trajectories in subjects with chronic obstructive pulmonary disease: a seven-year observation in clinical practice. J Patient-Reported Outcomes 2025; 9: 88.

36.

Chen

Gong

Bao

, et al. Pathways linking ICT use to chronic disease self-management among older adults with comorbidities in Shanghai. Healthcare 2025; 13: 1626.

37.

Shao

Yang

Chen

, et al. Determinants of digital health literacy among older adult patients with chronic diseases: a qualitative study. Front Public Health 2025; 13: 1568043.

38.

Liu

Guan

, et al. Neuroimaging insights into lung disease-related brain changes: from structure to function. Front Aging Neurosci 2025; 17: 1550319.

39.

Liang

Chen

, et al. Gray matter and cognitive alteration related to chronic obstructive pulmonary disease patients: combining ale meta-analysis and MACM analysis. Brain Imaging Behav 2025; 19: 204–217.

40.

Panda

Walsh

Anghel

, et al. Mechanisms and risk factors of cognitive impairment in chronic obstructive pulmonary disease. CHEST Pulmonary 2025: 100196.

41.

Peng

Wang

, et al. Latent profile analysis of dyspnea-related kinesiophobia in older adults with chronic obstructive pulmonary disease. Heart Lung 2025; 69: 241–247.

42.

Ouyang

, et al. Age-related differences in cognitive reappraisal: the role of cognitive function and implications for emotional regulation. Neuroscience 2025; 579: 312–320.

43.

Estacio

Whittle

Protheroe

. The digital divide: examining socio-demographic factors associated with health literacy, access and use of internet to seek health information. J Health Psychol 2019; 24: 1668–1675.

44.

Mooren

Smit

Engels

, et al. “When i am breathless now, i don’t have the fear that’s linked to it”: a case series on the potential of emdr to break the dyspnea-anxiety cycle in COPD. Bmc Pulm Med 2022; 22: 456.

45.

Brady

Mccloud

Higgins

, et al. The evolution of barriers and facilitators to using a COPD app among older adults: results from a pilot study. Front Digit Health 2025; 7: 1557590.

46.

Liu

Yin

. Loneliness and social isolation, mediating lifestyle factors, and incidence of COPD: a prospective cohort study. J Affect Disorders 2025; 383: 387–393.

47.

Wang

Zhou

Gao

, et al. Gamified working memory intervention enhances prefrontal neurocognitive plasticity during aging. Psychol Aging 2025; 40: 490–509.

48.

Poureslami

Kwan

Lam

, et al. Assessing the effect of culturally specific audiovisual educational interventions on attaining self-management skills for chronic obstructive pulmonary disease in Mandarin-and Cantonese-speaking patients: a randomized controlled trial. Int J Chronic Obstr 2016; 31: 1811–1822.

49.

Planert

Hildebrand

Machulska

, et al. Blended mobile-based interventions with integrated virtual reality exposure therapy for anxiety disorders: thematic analysis of patient perspectives. JMIR Hum Factors 2025; 12: e60957.

50.

Sohn

Rochester

Oluyase

. Features of COPD that lead to stigmatisation and its consequences: a framework synthesis. Copd 2025; 22: 2476435.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB