Association between generalized anxiety disorder and health-related quality of life

Abstract

Background

The objective of this study was to examine the association between generalized anxiety disorder (GAD) and health-related quality of life (HRQOL) in the general Korean population, and whether this relationship differed by multimorbidity profiles and household income.

Methods

We included 5,542 participants aged 19 years or older from the data of the 2021 Korean National Health and Nutrition Examination Survey. The GAD was assessed using the Generalized Anxiety Disorder 7-item scale (GAD-7). Participants with a GAD-7 score ≥ 10 were classified as having GAD. HRQOL was measured using the Health-related Quality of Life Instrument with 8 Items (HINT-8). Multimorbidity profiles were derived using latent class analysis. Survey-weighted regression models were used to evaluate associations and interaction effects.

Results

The weighted prevalence of GAD was 4.23%, with a mean GAD-7 score of 1.99. The mean HINT-8 index was 0.80. Higher GAD-7 scores and the presence of GAD were consistently associated with lower HINT-8 index values. GAD was associated with increased odds of problems across all HINT-8 domains, with particularly strong associations for depression (odds ratio [OR] = 34.43) and moderate associations for memory (OR = 4.88) and sleep (OR = 4.68). Three multimorbidity profiles (Healthy, Geriatric Disease, and Cardiometabolic Disease groups) were identified, and the association between GAD-7 scores and HRQOL differed by multimorbidity profile and household income, with similar patterns observed for GAD status.

Conclusions

Higher GAD-7 scores and GAD were associated with substantially poorer HRQOL across both physical and psychological domains. The strength of this relationship varied according to multimorbidity profile and socioeconomic context.

Background

Generalized anxiety disorder (GAD) is a pervasive mental health condition with a global prevalence of 3.7% [1], and its prevalence has been increasing over time [2]. Individuals suffering from GAD are more likely to visit emergency medical services, experience prolonged hospitalizations, and incur significantly higher annual medical costs [3]. A recent meta-analysis reported that direct medical costs for the GAD population are 2.6 times greater than for those without GAD [4], underscoring how its widespread prevalence places a massive financial burden on the overall national healthcare expenditure [5]. Furthermore, GAD is strongly associated with elevated risks of suicidal ideation and suicide attempts [6, 7]. Its symptoms adversely affect various aspects of individual health, including burden of comorbidity, productivity, and health-related quality of life (HRQOL) [8]. In the United States, a study focusing on older participants [9] found that the scores of two quality of life measures, Quality of Life Inventory (QOLI) and Life Satisfaction Index-Z (LSI-Z), were negatively associated with anxiety disorders measured by the Spielberger State-Trait Anxiety Inventory (STAI). A European study focusing on the general population across five countries [10] found that patients with GAD scored lower on the HRQOL measures, including Mental Component Summary (MCS) and Physical Component Summary (PCS), regardless of the presence of comorbidities. Similarly, a Canadian study that focused on the general population [11] found that individuals with GAD had lower life satisfaction and perceptions of well-being. A study conducted in China that focused on patients at mental health centers [12] has found that individuals with GAD scored lower on the quality of life measure, the World Health Organization Quality of Life-BREF (WHOQOL-BREF).

While the negative association between GAD and HRQOL is well documented [9–11], existing literature shows significant knowledge gaps regarding this relationship across different geographic areas and methodologies. Most studies have been heavily skewed toward Western populations, and little is known about the potential moderators of this association. To our knowledge, only two studies have explored how GAD is linked to HRQOL within Asian contexts [12, 13]; however, the generalizability of these findings is constrained, as they selectively focused on localized urban populations or psychiatric outpatients.

Although the clinical implications of GAD are increasingly complicated by multimorbidity—the coexistence of two or more chronic conditions within an individual—most studies have treated multimorbidity as a confounder. Diseases do not operate in isolation over the life course; rather, they interact in complex ways that can deteriorate an individual’s HRQOL and increase the burden of health management. In addition, a large share of the detrimental impacts of GAD and multimorbidity on HRQOL are profoundly shaped by structural socioeconomic inequalities. According to the intersectionality framework, health outcomes are not the sum of individual diseases and psychological distress; they are heavily moderated by economic circumstances such as employment status and household income. For instance, low-income individuals with psychological anxiety are likely to face the synergistic burden of chronic diseases and often experience a greater decline in their quality of life. Conversely, higher economic status may serve as a crucial buffering mechanism against adverse health outcomes caused by chronic diseases, preventing psychological distress from translating into severe problems that affect well-being. Despite the importance of comorbidity for understanding the mechanisms linking GAD and HRQOL, the complex intersectional dynamics among anxiety, comorbidity profiles, and the economic environment remain largely unexplored in nationally representative cohorts.

Therefore, this study aimed to assess the intersectional associations of GAD, latent classes of comorbidity status, and household income with HRQOL. Using nationally representative data from the Korea National Health and Nutrition Examination Survey (KNHANES), we sought to provide comprehensive, population-level evidence to guide targeted and integrated public health interventions.

Methods

We used data from the 2021 KNHANES, which was conducted by the Korea Centers for Disease Control and Prevention in 2021. The survey used a stratified, multistage probability sampling method based on age, sex, and geographical area. The survey sampled 25 households from 192 regions of Korea, ensuring a comprehensive representation of the Korean population. The KNHANES evaluated overall health, including mental health, through self-administered surveys and interviews with trained medical investigators. KNHANES data have been widely used in epidemiological studies examining chronic diseases, HRQOL, and mental health outcomes in Korea [14–17]. We obtained the 2021 KNHANES data from the KNHANES website. This study included 5,952 participants aged 19 years or older. Among them, 5,542 individuals (2,449 men and 3,093 women) were included in the analysis after excluding 410 participants who answered “don’t know” or provided “no response” to questions about GAD and HRQOL. The University of Seoul Institutional Review Board exempted this study from review as it was conducted with de-identified data for public welfare (IRB No. 2023-11-006).

Confounders

We utilized data on age, sex, smoking status, alcohol intake, exercise frequency, household income, education level, marital status, and physical comorbidities. Smoking status was categorized based on whether participants currently smoked. Current smokers were defined as those who had smoked more than five packs of cigarettes (100 cigarettes) in their lifetime and currently smoked. Alcohol intake was classified based on current consumption. Current drinkers were defined as those who consumed alcohol at least once per month in the past year. Exercise frequency was classified based on whether participants walked for at least 30 min at least once during the previous week. Household income was dichotomized into lower and higher groups based on the median household income for the study year. Education level was categorized into less than 10 years and 10 years or more, and marital status was classified based on the presence of a spouse. Depression was not included as a confounder, as it is conceptually incorporated into the depression dimension of the Health-related Quality of Life Instrument with 8 Items (HINT-8), and its inclusion could lead to overadjustment.

To adjust for the confounding effects of underlying health status, we employed the modified Elixhauser comorbidity index (ECI) based on the van Walraven weighted algorithm [18, 19]. The ECI was derived from physician-diagnosed chronic conditions available in the KNHANES database, including stroke, cardiovascular diseases (myocardial infarction or angina), chronic pulmonary disease (asthma), various types of cancer, hypertension, diabetes, and thyroid disease. Each condition was assigned a specific weight according to the van Walraven point system to calculate a cumulative score representing the comorbidity burden. The weighted ECI scores were categorized into three groups: 0 (no comorbidity), 1–7 (low-to-moderate burden), and ≥ 8 (high burden). A score of 8 was set as the cut-off to identify a high comorbidity burden, reflecting the weight assigned to severe conditions such as solid tumors. We also used each factor of underlying health status to construct latent classes of comorbidity that categorize the burden of multimorbidity based on the following conditions: hypertension, dyslipidemia, diabetes mellitus, arthritis, osteoporosis, and asthma.

Independent variables

We measured GAD using the Generalized Anxiety Disorder 7-item scale (GAD-7), which is one of the most commonly used tools for assessing anxiety [20, 21]. The GAD-7 assesses how often an individual has experienced each of the seven main symptoms of anxiety over the past two weeks. The response options included “not at all,” “several days,” “more than half the days,” and “nearly every day,” each scored on a scale of 0 to 3, resulting in a total score ranging from 0 to 21. Anxiety severity was categorized based on GAD-7 scores as follows: 0–4 for minimal (G1), 5–9 for mild (G2), 10–14 for moderate (G3), and 15–21 for severe anxiety (G4) [21]. A score of 10 was set as the cut-off to identify the presence of GAD [21, 22]. The Korean version of the GAD-7 has demonstrated good psychometric properties, including high internal consistency, test–retest reliability, and convergent validity in Korean populations [23–25].

Dependent variables

HRQOL was measured using the HINT-8, a tool customized specifically for Koreans [26]. Similar to widely used instruments such as the EuroQol five-dimension questionnaire (EQ-5D) and WHOQOL-BREF [27, 28], the HINT-8 assesses multiple domains of HRQOL. The HINT-8 comprises the following eight components: stair climbing, pain, energy, work, depression, memory, sleep, and happiness. The “stair climbing” item assesses the difficulty in climbing stairs; “pain” indicates physical discomfort; “energy” indicates the frequency of feeling energetic; “work” assesses the difficulty in performing tasks; “depression” measures the frequency of feeling depressed; “memory” evaluates the difficulty in remembering; “sleep” assesses the difficulty in sleeping; and “happiness” measures the frequency of feeling happy. Participants evaluated their condition over the past week for each component, with responses categorized into four levels according to the severity of the problem, ranging from Level 1 (no issues) to Level 4 (severe issues). The HINT-8 index was calculated using a specific formula [29], with index values ranging from 0.132 to 1. A value closer to 1 indicates a better quality of life, whereas a value closer to 0 indicates a poorer quality of life. The HINT-8 has demonstrated reliability and validity in Korean populations, including evidence of internal consistency and construct validity in both general and clinical samples [30, 31].

Statistical analysis

All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). We performed multivariable linear regression analyses, adjusting for confounders, to examine the associations of GAD-7 scores and GAD with the HINT-8 index. We examined trends in the proportions of reported problems across anxiety severity levels for each HINT-8 dimension and performed multivariable logistic regression analyses to estimate adjusted odds ratios for reporting problems in each HINT-8 dimension, comparing participants with and without GAD.

To identify unique subpopulations based on individuals’ concurrent chronic disease conditions, we performed a latent class analysis (LCA). To determine the optimal number of classes, we evaluated goodness-of-fit metrics including the Akaike information criterion (AIC) and Bayesian information criterion (BIC); clinical and theoretical meanings of each class were additionally assessed to label the classes. To examine whether the association between GAD-7 scores and the HINT-8 index differed according to multimorbidity profiles and household income, interaction terms among GAD-7 scores, multimorbidity class, and household income were included in the regression models.

Results

Of the 5,542 participants included in the analysis, 254 (weighted prevalence: 4.2%) screened positive for GAD (Table 1). Participants with GAD were younger and more likely to be female. They were less likely to engage in walking exercise, participate in economic activity, or have a spouse. Physician-diagnosed depression was more common in this group, and educational attainment differed between the groups. Smoking status, alcohol consumption, household income, and ECI score distribution were similar between the groups.

Table 1.

Characteristics of study participants according to GAD

	No GAD	GAD	P
n (weighted %)	5,288 (95.8)	254 (4.23)
Age, years	48.6 (0.45)	46.8 (1.31)	< 0.001
Male, %	50.8 (0.70)	33.5 (3.26)	< 0.001
Lifestyle variable, %
Current smoking	18.1 (0.78)	23.8 (3.42)	0.067
Alcohol intake	53.8 (1.02)	52.6 (3.85)	0.768
Exercise (≥ 30 min walking/week)	40.5 (1.09)	30.8 (3.67)	0.017
Socioeconomic status variable, %
Economic activity status	64.0 (0.91)	52.7 (3.66)	0.002
Lower income a	63.8 (1.54)	58.4 (3.83)	0.136
Education (< 10 years)	19.6 (1.10)	22.5 (3.02)	0.572
Spouse	74.7 (0.98)	68.6 (3.35)	0.062
Comorbidity burden by ECI b, %			0.338
0 (no comorbidity)	94.2 (0.36)	94.1 (1.29)
1–7 (low-to-moderate burden)	3.7 (0.29)	4.7 (1.14)
≥ 8 (high burden)	2.0 (0.20)	1.2 (0.54)
Mean of HINT-8 index	0.8150 (0.00168)	0.6586 (0.0084)	< 0.001
Positive responses of HINT-8 components c, %			< 0.001
Stair climbing	62.8 (0.88)	34.3 (3.82)
Pain	42.0 (0.90)	17.2 (2.85)
Energy	32.4 (0.71)	5.1 (1.29)
Work	57.1 (1.00)	12.6 (2.25)
Depression	55.0 (0.90)	3.2 (1.11)
Memory	48.5 (0.87)	18.6 (3.03)
Sleep	51.6 (0.82)	16.8 (3.13)
Happiness	25.2 (0.68)	1.7 (0.76)

Data are presented as n (weighted %), mean (SE), or percentage (SE)

^a Defined as household income below the median value for the study year

^b Defined as the ECI calculated using the van Walraven weighted method, excluding depression, and categorized as 0, 1–7, and ≥ 8 points

^c Defined as responses corresponding to the most favorable health state for each component (e.g., no difficulty, not at all depressed, or always happy)

Higher GAD-7 scores and the presence of GAD were both significantly associated with lower HINT-8 index scores (Table 2). The proportion of participants reporting the most favorable health states in stair climbing, pain, work, sleep, and happiness significantly decreased as anxiety severity levels increased (all p for trend < 0.001, Fig. 1). After adjusting for confounders, participants with GAD had 34.4- and 20.5-fold higher odds of reporting feelings of depression and unhappiness, respectively, compared with those without GAD (Table 3). Participants with GAD also had 3.4- to 10.4-fold higher odds of reporting problems in the remaining HINT-8 components.

Table 2.

Associations of GAD-7 scores and GAD with the HINT-8 index

Independent variable	GAD-7 score		GAD
Model	B (SE)	p	B (SE)	p
1	−0.0138 (0.00056)	< 0.001	−0.1547 (0.00815)	< 0.001
2	−0.0130 (0.00052)	< 0.001	−0.1410 (0.00697)	< 0.001
3	−0.0128 (0.00052)	< 0.001	−0.1406 (0.00687)	< 0.001

Model 1: adjusted for age and sex

Model 2: adjusted for age, sex, smoking status, alcohol intake, exercise frequency, economic activity status, household income, educational level, and marital status

Model 3: adjusted for age, sex, smoking status, alcohol intake, exercise frequency, economic activity status, household income, educational level, marital status, and ECI, calculated using the van Walraven method with depression excluded

Fig. 1.

Proportion of participants reporting any problems across the HINT-8 domains by anxiety severity level (all p for trend < 0.001). G1, G2, G3, and G4 indicate minimal, mild, moderate, and severe anxiety levels, respectively. Error bars represent SE

Table 3.

Odds ratios for problems in each HINT-8 component according to GAD status ^a

	Components of HINT-8
	Stair climbing	Pain	Energy	Work	Depression	Memory	Sleep	Happiness
Model 1	3.88 (2.74, 5.49)	3.60 (2.42, 5.37)	8.25 (4.84, 14.07)	10.45 (6.72, 16.25)	34.41 (16.83, 70.35)	4.55 (2.97, 6.99)	5.14 (3.27, 8.07)	19.13 (7.68, 47.62)
Model 2	3.62 (2.51, 5.21)	3.41 (2.24, 5.19)	7.52 (4.41, 12.82)	10.40 (6.59, 16.43)	34.53 (16.09, 74.12)	4.89 (3.12, 7.66)	4.70 (2.98, 7.39)	20.52 (7.41, 56.86)
Model 3	3.60 (2.50, 5.20)	3.40 (2.23, 5.17)	7.46 (4.37, 12.75)	10.40 (6.59, 16.43)	34.43 (16.05, 73.88)	4.88 (3.11, 7.65)	4.68 (2.97, 7.38)	20.50 (7.39, 56.83)

^a Odds ratios represent the likelihood of reporting problems (vs. no problems) in each component

Model 1: adjusted for age and sex

Model 2: adjusted for age, sex, smoking status, alcohol intake, exercise frequency, economic activity status, household income, educational level, and marital status

Model 3: adjusted for age, sex, smoking status, alcohol intake, exercise frequency, economic activity status, household income, educational level, marital status, and ECI, calculated using the van Walraven method with depression excluded

According to LCA model fit indices, the 3-class model was selected as it demonstrated the lowest BIC (21,718.36) while providing the most clinically interpretable subgroups (Table 4). Based on their distinct maximum posterior probabilities and comorbidity profiles, these three groups were designated as the Healthy group (Class 1: 65.7% of the sample, characterized by a generally low probability of all assessed diseases), the Geriatric Disease group (Class 2: 12.0% of the sample, showing distinctively higher probabilities for arthritis [62.4%] and osteoporosis [40.9%]), and the Cardiometabolic Disease group (Class 3: 22.3% of the sample, predominantly featuring high probabilities of hypertension [69.1%], dyslipidemia [59.8%], and diabetes [36.6%]) (Tables 5 and 6).

Table 4.

Model-fit indices for latent class models

Number of classes	Log-likelihood	df	AIC	BIC
2	−10844.4	13	21714.72	21800.78
3	−10,773	20	21585.96	21718.36
4	−10,761	27	21576.04	21754.77

Table 5.

Prevalence of latent classes and conditional probabilities

Characteristics	Class 1	Class 2	Class 3
Class marginal probabilities, %	65.7	12.0	22.3
Item response probabilities, %
Hypertension	7.7	52.8	69.1
Dyslipidemia	3.0	45.5	59.8
Diabetes mellitus	1.8	19.0	36.6
Arthritis	2.7	62.4	13.0
Osteoporosis	1.4	40.9	5.1
Asthma	0.8	6.8	2.5

Table 6.

Characteristics of study participants according to multimorbidity latent classes

	Healthy group	Geriatric Disease group	Cardiometabolic Disease group	p
n (weighted %)	3,979 (71.8)	547 (9.9)	1,016 (18.3)
Age, years	44.4 (0.39)	69.0 (0.46)	60.8 (0.48)	< 0.001
Male, %	52.1 (0.80)	15.7 (2.04)	54.9 (1.57)	< 0.001
Lifestyle variable, %
Current smoking	19.8 (0.81)	7.3 (1.64)	16.2 (1.51)	< 0.001
Alcohol intake	57.9 (1.05)	23.5 (2.31)	45.5 (1.85)	< 0.001
Exercise (≥ 30 min walking/week)	39.3 (1.18)	43.5 (2.68)	42.9 (1.84)	0.078
Socioeconomic status variable, %
Economic activity status	67.2 (0.99)	39.4 (3.05)	55.5 (1.98)	< 0.001
Lower income a	32.1 (1.53)	67.9 (2.67)	45.1 (2.26)	< 0.001
Education (< 10 years)	12.6 (0.83)	70.2 (2.49)	33.2 (2.03)	< 0.001
Spouse	68.5 (1.13)	98.0 (0.70)	94.7 (1.00)	< 0.001
Comorbidity burden by ECI b, %
0 (no comorbidity)	96.8 (0.30)	82.8 (1.95)	86.5 (1.28)	< 0.001
1–7 (low-to-moderate burden)	1.7 (0.22)	13.0 (1.74)	10.1 (1.14)	< 0.001
≥ 8 (high burden)	1.5 (0.19)	4.2 (0.94)	3.4 (0.72)	< 0.001
Mean of HINT-8 index	0.82 (0.002)	0.72 (0.006)	0.78 (0.004)	< 0.001
Positive responses of HINT-8 components c, %
Stair climbing	68.4 (0.84)	19.0 (1.95)	45.6 (1.88)	< 0.001
Pain	44.7 (0.99)	13.3 (1.71)	34.0 (1.62)	< 0.001
Energy	31.6 (0.83)	23.6 (1.87)	32.8 (1.91)	0.007
Work	59.6 (1.06)	27.9 (2.18)	44.7 (1.94)	< 0.001
Depression	53.2 (1.05)	42.3 (2.44)	55.2 (1.97)	0.0003
Memory	50.6 (0.97)	27.0 (2.23)	39.2 (1.72)	< 0.001
Sleep	52.3 (0.95)	35.0 (2.21)	45.4 (2.05)	< 0.001
Happiness	24.9 (0.77)	18.2 (1.73)	23.1 (1.59)	0.010

Data are presented as n (weighted %), mean (SE), or percentage (SE)

^a Defined as household income below the median value for the study year

^b Defined as the ECI calculated using the van Walraven weighting method, excluding depression, and categorized into 0, 1–7, and ≥ 8 points

^c Defined as responses corresponding to the most favorable health state for each component (e.g., no difficulty, not at all depressed, or always happy)

Across the entire sample, a clear inverse relationship between GAD-7 scores and the HINT-8 index was observed uniformly across all comorbidity classes (p for two-way interaction = 0.48, Fig. 2). Notably, the Healthy group maintained the highest baseline HRQOL at any given GAD-7 score, followed by the Cardiometabolic Disease group, while the Geriatric Disease group consistently exhibited the lowest HINT-8 index. When we stratified the predicted probabilities by household economic status, however, marked differences in the slopes emerged across the three comorbidity groups (p for three-way interaction = 0.023). Among individuals in lower-income households, the decline in HRQOL associated with increasing GAD-7 scores was notably steeper among individuals in the Cardiometabolic Disease group (Fig. 3). This disparity in the slope between the Healthy and Cardiometabolic Disease groups was attenuated among individuals with higher household income. In other words, the decline in the HINT-8 index with increasing GAD-7 scores among individuals in the Cardiometabolic Diseases was less steep in higher-income groups, suggesting that greater financial resources may mitigate the impact of health-related challenges. Similar patterns were observed when GAD status was used instead of GAD-7 scores.

Fig. 2.

Predicted HINT-8 index across GAD-7 scores by multimorbidity latent class groups, adjusted for confounders (p for two-way interaction = 0.48)

Fig. 3.

Predicted HINT-8 index across GAD-7 scores by multimorbidity latent class group, adjusted for confounders and stratified by household income (p for three-way interaction = 0.023). (A) lower income, (B) higher income

Discussion

This study showed that GAD was associated with poorer HRQOL. GAD was significantly associated with both the physical (stair climbing, pain, and work) and psychological (depression, memory, energy, happiness, and sleep) aspects of HRQOL. These findings are consistent with prior studies conducted on the general population across five European countries and among patients at mental health centers in China [10, 12].

It is plausible that the association between GAD and the physical domains of HRQOL is partly explained by activation of the sympathetic nervous system, which can directly result in physical fatigue and lack of energy [32]. In addition, individuals with GAD may be treated with psychotropic medications, whose side effects, including insomnia and mood disturbances, may further exacerbate reductions in physical activities. Anxiety-related activation of this hormone system may also disrupt serotonin and norepinephrine balance, thereby negatively influencing the psychological domains of HRQOL [33, 34]. More broadly, such neuroendocrine dysregulation may undermine mental stability and increase vulnerability to GAD and depression. Our findings showed that GAD was highly associated with depression within the psychological domains. This implies that depression and GAD share a similar etiological mechanism. They are both affected by imbalances in brain neurotransmitters including serotonin [35, 36]. Additionally, both conditions are associated with functional and structural changes in specific brain regions, such as the hippocampus, prefrontal cortex, and amygdala [37, 38]. These changes can contribute to the interaction between the two conditions, increasing psychological burdens.

We next examined the interactions of comorbidity and socioeconomic resources with GAD. We found parallel slopes for the relationships between HRQOL and GAD-7 scores across all comorbidity classes. While the presence and type of comorbidities—such as cardiometabolic or geriatric diseases—significantly lowered the baseline HRQOL (represented by different intercepts), the marginal negative impact of escalating anxiety remained constant across these groups. This finding implies that the negative impact of psychological distress indicated by GAD is neither accelerated nor mitigated by the presence of underlying health conditions. Thus, interventions targeting anxiety management are likely to yield general improvements in quality of life across all patient populations, regardless of their baseline health conditions.

When we added the interaction terms with household income, however, we found significant differences in the slopes of the relationships across comorbidity profiles. This evidence suggests that higher economic resources may exert a protective buffering effect that specifically mitigates the sharp, disproportionate decline in HRQOL observed in the lower-income Cardiometabolic Disease group. This specific vulnerability identified in the lower-income group underscores a vital public health implication. It demonstrates that the detrimental synergy between anxiety and physical comorbidity is not strictly a biological phenomenon, but one strongly moderated by socioeconomic disparity. This pattern may be explained by several factors. First, lower-income participants may have more limited access to timely care, rehabilitation services, and other supportive resources because of financial constraints and competing life demands, making it more difficult to manage both anxiety symptoms and cardiometabolic conditions [39]. Second, ongoing financial stress may exacerbate anxiety and increase the psychological burden of living with multimorbidity [40, 41]. Third, many cardiometabolic conditions require sustained self-management, including medication adherence, lifestyle modification, and symptom monitoring [42]. Anxiety symptoms may disrupt these daily self-management behaviors, while limited resources in low-income settings may further hinder effective disease management [42, 43]. Therefore, clinical interventions aiming to improve HRQOL should not rely solely on generalized anxiety management; rather, they must prioritize targeted, multidisciplinary support—combining mental health care with socioeconomic assistance—specifically tailored for low-income patients burdened with cardiometabolic diseases.

To our knowledge, this is the first population-based study in Asia to examine the relationship between GAD and HRQOL while simultaneously considering multimorbidity patterns and socioeconomic context. By applying LCA to identify multimorbidity profiles, our study moves beyond treating multimorbidity as a simple count or confounder and instead captures the complex clustering of cardiometabolic conditions. In addition, by incorporating household income, we were able to explore the intersectional interplay between psychological symptoms, cardiometabolic disease patterns, and socioeconomic conditions in shaping HRQOL. Finally, this study used the HINT-8, a culturally adapted HRQOL instrument developed and validated in Korea, addressing limitations of previous research that relied primarily on instruments developed in Western contexts [30, 31, 44, 45].

However, this study had a few limitations. First, as this was a cross-sectional study, a causal relationship between GAD and HRQOL could not be established. Second, factors influencing GAD, such as life events and medication history [46, 47], were neither investigated nor included as covariates. Lastly, our modified ECI cannot rule out the possibility of reporting bias or underestimation, given that it was derived from physician-diagnosed conditions collected through self-reported interviews rather than administrative claims data or medical records, and did not incorporate all 31 original conditions due to the inherent constraints of the KNHANES database.

This study showed an inverse relationship between GAD and HRQOL in the general Korean population. This association may be explained by multiple underlying pathways and other potential factors contributing to the development of GAD. We also found complex interactions among GAD, comorbidity, and the socioeconomic environment in affecting health-related quality of life. To clarify these multi-dimensional relationships, additional research is required to confirm the causal links between GAD and HRQOL using longitudinal datasets.

Abbreviations

Abbreviation

Full description

GAD

Generalized anxiety disorder

HRQOL

Health-related quality of life

WHOQOL-BREF

World Health Organization Quality of Life-BREF

KNHANES

Korea National Health and Nutrition Examination Survey

HINT-8

Health-related Quality of Life Instrument with 8 Items

ECI

Elixhauser comorbidity index

GAD-7

Generalized Anxiety Disorder 7-item scale

LCA

Latent class analysis

AIC

Akaike information criterion

BIC

Bayesian information criterion

Author contributions

Jiseon Kim: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft.Yeonjin Lee: Formal analysis, Methodology, Software, Visualization, Writing – review & editing.Sangshin Park: Conceptualization, Methodology, Project administration, Supervision, Validation, Writing – review & editing.

Funding

No funding was received for this study.

Data availability

The data used in this study are publicly available from the Korea National Health and Nutrition Examination Survey (KNHANES) through the official Korea Disease Control and Prevention Agency website.

Declarations

Ethics approval and consent to participate

The University of Seoul Institutional Review Board deemed that this study was conducted with de-identified data for public welfare, and no ethics board approval was required (IRB No.2023-11-006). Informed consent had been obtained from participants by the KNHANES during the original data collection. The survey was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.