Typologies of Symptoms and Functions Among Patients With Chronic Kidney Disease and Population Characteristics: A Latent Profile Analysis

ABSTRACT

Aims

To explore the latent classification of symptom distress and functional impairments in Chinese patients with chronic kidney disease (CKD) and analyse the influencing factors of different latent classes.

Design

A cross‐sectional survey study.

Methods

Data were collected from a public tertiary hospital in Shanghai, China, between December 2023 and January 2024. Data collection was performed with the use of the Patient‐Reported Outcomes Measurement Information System −29 Profile V2.1. Latent Profile Analysis and a multinomial logistic regression model were employed to identify subgroups among patients.

Results

The results indicated three distinct subgroups of symptom distress and functional impairments among patients: Profile 1—‘moderate to severe symptom distress and moderate to severe functional impairments’ (29, 13.9%), Profile 2—‘low symptom distress and no functional impairments’ (134, 64.1%) and Profile 3—‘no symptom distress and no functional impairments’ (46, 22.0%). Various socio‐demographic factors, such as age, payment methods, daily care by myself, anaemia, diabetes and disturbance of phosphorus metabolism, demonstrated significant associations of these factors with membership in the latent profiles (p < 0.05).

Conclusion

This study identified three symptoms and functional impairment profiles in patients with CKD and the predictors of the profile membership. These findings highlight the need for personalised interventions for patients with CKD.

Implications for the Profession and/or Patient Care

Early identification of high‐risk patients with CKD who experience moderate to severe symptom distress and functional impairments is crucial. Providing targeted interventions, including financial and daily care support, as well as managing comorbidities such as diabetes, anaemia and disturbances in phosphorus metabolism, is essential for achieving effective outcomes.

Reporting Method

This study followed the STROBE statement of cross‐sectional studies.

Patient or Public Contribution

The study was conducted by patients, healthcare professionals, and the research team.

1. Introduction

Chronic Kidney Disease (CKD), characterised by a comprehensive classification encompassing a spectrum of conditions marked by kidney impairment and/or a reduction in Glomerular Filtration Rate (GFR) (falling below 60 mL/min/1.73m²) over 3 months, includes conditions such as diabetic nephropathy, hypertensive nephrosclerosis, polycystic kidney disease, glomerulonephritis and nephrotic syndrome (Kalantar‐Zadeh et al. 2021). CKD is classified into five stages based on kidney function, which is assessed using the GFR. Stage 1 is characterised by normal kidney function (GFR ≥ 90 mL/min/1.73m²) but with evidence of kidney damage. Stage 2 (GFR 60–89 mL/min/1.73m²) reflects a mild reduction in kidney function. Stage 3 is further divided into 3a (GFR 45–59 mL/min/1.73m²) and 3b (GFR 30–44 mL/min/1.73m²), indicating moderate kidney damage. Stage 4 (GFR 15–29 mL/min/1.73m²) denotes severe reduction in kidney function, while Stage 5 (GFR < 15 mL/min/1.73m²) represents kidney failure, often necessitating dialysis or a kidney transplant (Inker et al. 2014).

CKD is the third fastest‐growing cause of death globally and the only non‐communicable disease with a rising age‐standardised mortality rate year‐over‐year. It is projected that CKD will become the fifth leading cause of death worldwide in 2040. Currently, an estimated 843.6 million people globally are affected by CKD (Bello et al. 2024), accounting for approximately 13% of the global population. A literature review covering 161 countries found the median global prevalence of CKD to be 9.5%, with the highest rate of 12.8% in Eastern and Central Europe (Bello et al. 2024). Moreover, the global prevalence of CKD continues to rise rapidly, typically affecting younger demographics. In China, the prevalence of CKD stands at 10.8%, with a rate as high as 18.3% in the southwestern regions (Shi et al. 2024). Additionally, studies indicate that the number of patients with CKD in China is expected to increase by an average of 2.6 million (1.6%) each year from 2020 to 2025, and the total economic burden of CKD in China is projected to increase by an average of $3.1 billion annually (Shi et al. 2024).

Given the substantial and growing financial pressure, China's multilevel medical security plays a critical role in mitigating medical costs for patients with CKD. The medical security system, with the basic medical insurance (BMI) as its pillar, provides coverage for CKD‐related treatments including dialysis and transplantation. However, reimbursement rates exhibit significant variations depending on regional healthcare policies, hospital grade and insurance type, with employed individuals covered under Employee BMI (EBMI) and non‐working residents under Resident BMI (RBMI) (Yi 2021; Zheng et al. 2023). Medical insurance has played a crucial role in alleviating the financial burden on patients with CKD and significantly improves healthcare accessibility.

Symptom distress in patients with CKD encompasses a range of physical and psychological challenges that significantly impact their quality of life. Research has shown that symptom distress is closely associated with the deterioration of both physical and mental health in patients with CKD (Tang et al. 2017). Patients with CKD experience symptoms of both physical aspects (e.g., sleep disturbances, fatigue, reduced appetite, swelling in the limbs, sarcopenia and muscle cramps) and psychological aspects (e.g., depression, anxiety and decreased mental clarity) (Morishita et al. 2017). The burden of managing a chronic illness can lead to anxiety, depression and social isolation (Iovino et al. 2023). Research indicated that patients with CKD experienced a worsening of emotional state due to feelings of disappointment and depression about their illness, and this state leads to fatigue, decreased vitality and even declined sleep quality (Gregg et al. 2021). This, in turn, further disrupted the patients' social functions caused more unstable emotional state, and exacerbated the patients' biochemical imbalance (Picariello et al. 2017). As a result, a vicious cycle of maintenance was created.

Research indicated that a decline in physical function is one of the most common impacts of CKD (Flythe et al. 2021). Physical function encompasses an individual's ability to perform activities that require physical capability, ranging from basic daily care by myself to more intense activities (Painter et al. 1999). Previous studies have highlighted the physical function in patients with CKD, focusing mainly on its impact on patients' ability to engage in exercise (Weber et al. 2021; Jhamb et al. 2016). The 2012 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines recommended that patients engage in at least 30 min of physical activity tailored to their cardiovascular health and tolerance five times a week (Stevens and Levin 2013). Moreover, studies have confirmed that nearly all the patients who reported limitations in physical function also experienced negative impacts on their ability to participate in activities, including routine household chores and social interactions (Palmer et al. 2024). Reduced social interactions would lead to feelings of isolation and loneliness, exacerbating patients' psychological symptoms (Palmer et al. 2024).

CKD impacts a wide range of patients, from young individuals with genetic or glomerular diseases to elderly adults suffering from kidney‐affecting conditions like hypertension and diabetes. As a result, symptoms displayed by patient with CKD demonstrate significant variability across the population (Chong and Unruh 2017). Although some studies have used variable‐centred approaches to examine associations among variables such as symptoms and related factors, they have overlooked significant individual variations in these symptoms (Rosendal et al. 2013). The person‐centred approaches facilitate the identification of subpopulations characterised by heterogeneous features. Latent class models (LCMs) are statistical tools for building typologies based on observed variables and are especially helpful to identify subgroups within heterogeneous populations (Tein et al. 2013). Latent profile analysis enables researchers to categorise patients with CKD based on symptom distress and functional impairments, and it would be helpful for the development of tailored support interventions for each subgroup (Vermunt and Magidson 2002).

Therefore, based on different population characteristics, our study aimed to identify the subgroups characterised by symptom distress and functional impairments among patients with CKD, with population heterogeneity considered. Knowledge regarding the profiles and factors that predict the profile membership could provide valuable guidance for healthcare professionals to perform symptom management and functional improvement interventions for patients with CKD.

1.1. Theoretical Framework

The theory of unpleasant symptoms (TOUS) (Lenz et al. 1997) is a structurally complicated framework that would be a conceptual fit for chronic disease if family influence and perceived function were included (Silva‐Rodrigues et al. 2019). According to the TOUS, physiological, psychological and situational factors have impacts on symptom characteristics and their outcomes. Under the instruction of the TOUS, the physiological factors (patient's gender, age, stages of disease, diabetes, hypertension, hypourocrinia, proteinuria, anaemia, and disturbance of calcium, phosphorus, potassium and sodium metabolism), psychological factors (anxiety and depression) and situational factors (educational status, religion, marital status, owning a pet, employment status, average monthly household income, payment methods for medical expenses and daily care by myself) were collected. Performance outcomes, including symptom distress (anxiety, depression, fatigue, sleep disturbance, pain interference and pain intensity) and function (physical function (mobility), and ability to participate in social roles and activities) were assessed.

2. Methods

2.1. Design

This was a descriptive cross‐sectional study. This study was prepared and reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).

2.2. Participants

A convenience sample of patients with CKD was recruited from a public tertiary hospital in Shanghai, China. Patients diagnosed with CKD were eligible to participate if they were able to self‐report the health‐related outcomes and if they volunteered for the survey. The exclusion criteria were as follows: (1) self‐reported severe visual, hearing or verbal communication impairments, and (2) a history of cognitive impairment disorders (e.g., Alzheimer's disease, dementia) or other psychiatric illnesses, as documented in medical records. For the sample size of the Latent Class Model (LCM), based on past research experience, a simple LCM analysis usually should be conducted with a minimum sample size of 100–150 or more cases (Wang and Wang 2012). In this study, the sample size exceeds 200 cases and therefore is reasonable.

2.3. Data Collection

A cross‐sectional survey of patients with CKD was conducted in a hospital in Shanghai from December 2023 through January 2024. Approval was obtained from the institutional review board of The People's Liberation Army Naval Characteristic Medical Center (AF‐HEC‐064) before the study started, and then informed consent was obtained from patients with CKD.

The researchers selected and trained the research nurses. Thereafter, the research nurses selected participants who met the inclusion and exclusion criteria and first explained the purpose of the study to the participants, clarified that their participation was voluntary, anonymous and confidential, and informed participants that they could withdraw at any time. After obtaining their verbal consent, the research nurses then distributed the printed questionnaires and reviewed them with the participants. A total of 209 valid questionnaires were collected after eliminating 11 unqualified questionnaires with more than 50% missing items.

2.4. Measures

2.4.1. Demographic Questionnaire

The socio‐demographic and medical information was collected from patients. The information included: (1) patient's sociodemographic information: gender, age, educational status, religion, marital status, owning a pet, employment status, average monthly household income, payment methods for medical expenses and daily care by myself; (2) patient's medical information: stages of CKD (using the CKD‐EPI equation), diabetes, hypertension, hypourecinia, proteinuria, anaemia, and levels of calcium, phosphorus, potassium and sodium.

2.4.2. PROMIS

The Patient‐Reported Outcomes Measurement Information System (PROMIS) represents a collaborative initiative involving eight US research institutes and the US National Institutes of Health (NIH). The Patient‐Reported Outcomes Measurement Information System −29 Profile V2.1 (PROMIS‐29 V2.1), developed via qualitative and item response theory methods, is one of the most frequently employed PROMIS measures, and has already been translated into Chinese (Cai et al. 2021). PROMIS‐29 V2.1 includes 29 items across seven domains, with five symptom domains: anxiety, depression, fatigue, sleep disturbance, pain interference and pain intensity, and two function domains: physical function (mobility) and ability to participate in social roles and activities. A 5‐point Likert scale (range, 1–5) was used to measure symptom severity or frequency, with responses on the mobility subscale ranging from ‘with no trouble’ (5 points) to ‘not able to do’ (1 point). Two of the five items for sleep disturbance were from ‘very poor’ (5 points) to ‘very good’ (1 point) and from ‘never’ (5 points) to ‘almost always’ (1 point), respectively; the other subscales asked about the frequency of experiencing the respective health domain, ranging from ‘never’ (1 point) to ‘almost always’ (5 points). The single pain intensity item was scored separately, with response scale ranging from 0 (no pain) to 10 (worst pain imaginable). This item was neither analysed nor discussed since it is not applicable to LPA. Questions regarding physical function and ability to participate in social roles and activities did not provide a specific time frame. Regarding the other 5 domains, questions were asked concerning the past 7 days. Domain scores were obtained by summing the item scores for each domain. The raw score was translated into T‐scores according to the PROMIS scoring manual (Cai et al. 2021). According to PROMIS guidelines for cut points, higher symptom scores reflect more severe symptom distress, with cut points of no symptom distress (< 55), mild symptom distress (55–59), moderate symptom distress (60–69), and severe symptom distress (≥ 70) (Cai et al. 2021). Scores showing lower physical function (mobility) or decreased ability to participate in social roles and activities reflect worse functioning, with cut points of no impairment (> 45), mild impairment (41–45), moderate impairment (31–40) and severe impairment (≤ 30) (Cai et al. 2021).

The linguistic equivalence and cross‐cultural validity of the Chinese version of the PROMIS‐29 have been verified (Cai et al. 2021; Cai, Huang, et al. 2022; Liu et al. 2016). The PROMIS‐29 has been tested for validity and reliability in patients with cancer in China (Cronbach's α = 0.88 ~ 0.95) (Tang et al. 2019). In this study, Cronbach's α was 0.857 ~ 0.914 for the 7 domains, indicating satisfactory internal consistency.

2.5. Statistical Analysis

Data were analysed with the use of SPSS and Mplus. Descriptive analyses were conducted for the characteristics of patients with CKD and the PROMIS‐29 results. All missing data were handled via full information maximum likelihood (FIML) in Mplus. LPAs with one to four latent profiles were conducted to identify subgroups of symptoms and functions in patients with CKD. The multinomial logistic regression model was used to assess the association of the individual characteristics with the latent profile membership. LPA was estimated with the use of Mplus 8.0 for Mac, and other analyses were conducted with the use of SPSS 26.0. All tests of statistical significance were 2‐sided at α = 0.05.

To conduct the LPA, we used 200 sets of random start values as well as 50 initial stage iterations to ensure that the model estimates were not based on local maxima of the likelihood function. Model fit was evaluated via indices as recommended (Sinha et al. 2021): Akaike information criterion (AIC), Bayesian Information Criterion (BIC) and sample size adjusted BIC (ABIC). For selecting the optimal number of Classes, we performed tests, such as the Lo–Mendell–Rubin test (LMR), Lo–Mendell–Rubin adjusted LRT (aLMR), and bootstrapped likelihood ratio test (BLRT), to compare a model with k classes to one with k−1 classes (Dziak et al. 2014). Additionally, models were also evaluated via entropy, a measure of classification accuracy ranging between 0 and 1, with values above 0.80 indicating high discrimination across latent classes (Jichuan and Xiaoqian 2019). Lower indices (including AIC, BIC and ABIC), higher test values (including LMR and BLRT), and higher entropy values indicate better model fit.

3. Results

3.1. Participant Characteristics

Participants included 209 patients with CKD, whose sociodemographic and medical characteristics are described in Table 1.

TABLE 1.

Characteristics of patients with CKD (n = 209).

Variables	Categories	N	%
Gender	Male	123	58.9
	Female	86	41.1
Age (years old)	< 40	41	19.6
	40–60	54	25.8
	> 60	114	54.5
Educational status	Junior school and below	105	50.2
	High school	50	23.9
	Junior college or above	54	25.9
Religion	No	175	83.7
	Yes	34	16.3
Marital status	Married	191	91.4
	Others (divorced, widowed, or separated)	18	8.6
Owning a pet	Yes	34	16.3
	No	175	83.7
Employment status	Full‐time (≥ 40 h per week)	50	23.9
	Part‐time (< 40 h per week)	14	6.7
	Others	145	69.4
Average monthly household income	< 2000 (< $289)	39	18.7
	2000–5000 ($289–723)	67	32.1
	5000–8000 ($723–1157)	81	38.8
	> 8000 (> $1157)	22	10.4
Payment methods for medical expenses	Employee basic medical insurance	114	54.5
	Residents basic medical insurance	40	20.1
	Self‐paying	53	25.4
Daily care by myself	Yes	130	62.2
	No	79	37.8
Stages of disease	CKD stage 5	50	23.9
	CKD stage 4	40	19.2
	CKD stage 3	56	26.8
	CKD stage 2	35	16.7
	CKD stage 1	28	13.4
Diabetes	Yes	82	39.2
	No	127	60.8
Hypertension	Yes	173	82.8
	No	36	17.2
Hypourocrinia	Yes	50	23.9
	No	159	76.1
Proteinuria	Yes	152	72.7
	No	57	27.3
Anaemia	Yes	148	70.8
	No	61	29.2
Disturbance of calcium metabolism	Yes	103	49.3
	No	106	50.7
Disturbance of phosphorus metabolism	Yes	58	27.8
	No	151	72.2
Disturbance of potassium metabolism	Yes	29	13.9
	No	180	86.1
Disturbance of sodium metabolism	Yes	26	12.4
	No	183	87.6

3.2. Latent Profiles of Symptom Distress and Functional Impairments in Patients With CKD

Results from the one‐ to four‐profile LPA supported a 3‐profile solution. Model fit information for all models is presented in Table 2. Apparently, the 1‐profile solution had the worst model fit (largest AIC, BIC and ABIC) and was therefore rejected. The 3‐profile solution was better than the 2‐profile due to lower AIC, BIC and ABIC values, significant LMR and aLMR values, and slightly higher entropy. The participants' population was shown to be heterogeneous with respect to the seven domains of PROMIS‐29. The 4‐profile solution, despite having slightly lower AIC, BIC and ABIC values and higher entropy, was not statistically different from the three‐profile solution based on aLMR values. For the balance of model fit, model parsimony, and clinical interpretability of the results, the three‐profile solution was selected. The entropy value of 0.975 suggested a clear delineation of latent profiles.

TABLE 2.

Fit results of potential profile models for self‐reported outcomes in patients with CKD (n = 209).

Models	k	AIC	BIC	ABIC	Entropy	LMR	aLMR	BLRT	Relative class size
C1	14	11186.451	11233.244	11188.885	—	—	—	—	—
C2	22	10577.804	10651.336	10581.628	0.919	0.0002	0.0003	< 0.0001	0.60/0.40
C3	30	10300.429	10400.699	10305.643	0.975	0.0083	0.0091	< 0.0001	0.139/0.641/0.22
C4	38	10102.775	10229.784	10109.380	0.947	0.0499	0.0532	< 0.0001	0.22/0.10/0.31/0.37

The mean standardised scores for each domain of PROMIS‐29 with respect to the 3‐profile solution are presented in Figure 1 and Table 3. Profile 1 accounts for 13.9%, with scores of five dimensions of symptoms above the reference line and scores regarding the dimensions of function‐mobility and ability to participate in social roles and activities significantly below the reference line, and scores significantly below the reference line on the dimensions of function‐mobility and ability to participate in social roles and activities. Hence, Profile 1 was labelled as ‘moderate to severe symptom distress and moderate to severe functional impairments’. Likewise, Profile 2 accounted for 64.1%, with scores of five dimensions of symptoms approaching the reference line and scores of two functional dimensions slightly higher than the reference line. Profile 2 was thereby labelled as ‘low symptom distress and no functional impairments’. In Profile 3, patients with CKD (46, 22.0%) scored below the reference line in the five symptom dimensions and above the reference line in the two functional dimensions. Therefore, Profile 3 was labelled as ‘no symptom distress and no functional impairments’.

FIGURE 1.

Sores of seven domains of PROMIS‐29 in three profiles (n = 209).

TABLE 3.

Sores of seven domains of PROMIS‐29 (n = 209).

Domains	Profile 1	Profile 2	Profile 3	Critical value
Anxiety	71.623 ± 11.063	53.039 ± 8.366	43.748 ± 5.640	55
Depression	66.627 ± 11.428	51.201 ± 8.512	41.937 ± 3.232	55
Fatigue	67.493 ± 5.911	55.353 ± 3.926	34.630 ± 2.498	55
Sleep disturbance	63.840 ± 9.475	50.511 ± 7.633	39.885 ± 7.431	55
Pain influence	63.740 ± 10.159	51.412 ± 9.480	43.883 ± 5.410	55
Physical function‐mobility	27.290 ± 6.634	46.741 ± 7.782	55.387 ± 3.674	45
Ability to participate in social roles and activities	33.037 ± 6.853	50.358 ± 7.822	62.707 ± 3.259	45

3.3. Multivariate Analysis of Potential Profiles of Self‐Reported Outcomes in Patients With CKD

The results of the multivariate analysis of potential profiles are shown in Table 4. Compared with patients with CKD over 40 years old, those under 40 were more likely to be in Profile 3 rather than in Profile 2 (OR = 0.060, p < 0.01). Compared with patients whose payment methods for medical expenses were self‐paying, patients who had medical insurance were more likely to be in Profile 3 than in Profile 1 and Profile 2 (OR = 0.076, p < 0.05; OR = 0.161, p < 0.05). Compared with patients who had family members as caregivers, patients who provided daily care by themselves were more likely to be classified in Profile 3 than in Profile 1 (OR = 0.021, p < 0.01). Patients with diabetes were more likely to be classified in Profile 1 than in Profile 3 (OR = 25.929, p < 0.01). Additionally, patients with anaemia were more likely to be classified in Profile 1 and Profile 2 than in Profile 3 (OR = 6.934, p < 0.05; OR = 5.858, p < 0.01). Moreover, patients with disturbance of phosphorus metabolism were more likely to be classified in Profile 1 than in Profile 3 (OR = 10.029, p < 0.05).

TABLE 4.

Results of the multinomial logistic regression models (n = 209).

Variables	Profile 1							Profile 2
	β	SE	Wald χ 2	OR	95% CI		p	β	SE	Wald χ 2	OR	95% CI		p
					Lower limit	Upper limit						Lower limit	Upper limit
Age (years old)
< 40	−1.476	1.4	1.113	0.228	0.015	3.549	0.291	−2.822	0.766	13.559	0.06	0.013	0.267	0.000
Payment methods for medical expenses
Medical insurance	−2.574	1.096	5.518	0.076	0.009	0.653	0.019	−1.829	0.746	6.009	0.161	0.037	0.693	0.014
Self‐care	−3.858	1.132	11.607	0.021	0.002	0.194	0.001	−0.730	0.676	1.164	0.482	0.128	1.815	0.281
Diabetes	3.255	0.992	10.764	25.929	3.708	181.29	0.001	0.057	0.553	0.011	1.059	0.358	3.129	0.918
Anaemia	1.936	0.951	4.148	6.934	1.076	44.692	0.042	1.768	0.543	10.600	5.858	2.021	16.982	0.001
Disturbance of phosphorus metabolism	2.305	0.976	5.577	10.029	1.48	67.963	0.018	−0.608	0.731	0.692	0.544	0.130	2.282	0.406

4. Discussion

4.1. Symptom Distress and Functional Impairments Among Patients With CKD Exhibit Group Heterogeneity

The symptom distress and functional impairments among patients with CKD exhibited inter‐group heterogeneity and can be classified into three latent profiles: Profile 1‐ moderate to severe symptom distress and moderate to severe functional impairments (29, 13.9%), Profile 2‐ low symptom distress and no functional impairments (134, 64.1%), and Profile 3‐ no symptom distress and no functional impairments (46, 22.0%); most of (163, 78.0%) patients with CKD in this study experienced a spectrum of symptoms. In addition, the results in this study showed that the symptom distress of patients with CKD had a negative influence on functional impairments, a finding that was consistent with the findings of the previous study (Li et al. 2021). The moderate to severe symptom distress in patients with CKD had an impact on physical and social functional impairments, while mild symptom distress in patients with CKD had little impact on physical and social functions. Hence, managing both physiological and psychological symptoms of patients with CKD was of great importance, as it not only helped to improve patients' physical and social function, but also contributed to the improvement of patients' quality of life.

Patients in Profile 1 manifested the most severe symptoms, along with significant impairments in physical and social function, especially for the severe anxiety symptoms and mobility function impairment. Due to the physiological symptoms such as pain and fatigue caused by the disease itself, patients with CKD tended to reduce their physical activity and opportunities to socialise with others, thus negatively affecting their social function (Sluiter et al. 2024). The limited social function constituted a negative feedback loop at their psychological status. Research has shown that the weakening of social functions can exacerbate psychological symptoms such as anxiety and depression in patients (Picariello et al. 2017), which in turn can worsen physical symptoms, creating a vicious cycle and seriously affecting the overall recovery process and quality of life of patients with CKD (Almutary et al. 2017; Bonner et al. 2014).

4.2. Association Between Socio‐Demographic, Clinical Characteristics and Latent Profiles

4.2.1. The Age

Patients with CKD under 40 years old were more likely to be classified in Profile 3. Unlike CKD in younger individuals, CKD in older adults may result from multiple chronic pathological mechanisms that accelerate functional impairments before end‐stage CKD becomes apparent (Liu et al. 2021). Additionally, findings of this study contrasted significantly with the previous research in younger cohorts (Senanayake et al. 2017; Yapa et al. 2021; Liu et al. 2024), which showed heavier symptom distress in young patients with CKD, particularly in terms of frailty and psychological issues like anxiety and depression. The variations could be due to differences in the subjects included across different regions. For example, some studies include young subjects engaged in physically demanding jobs, which may exacerbate their symptom distress (Liu et al. 2024). Although this study found that patients with CKD under 40 tended to experience no symptom distress and no functional impairments, regular monitoring and early symptom identification are essential to prevent the worsening of symptoms in adults with CKD. Patients should also be involved in symptom management plans tailored to their preferences and needs (Yapa et al. 2020).

4.2.2. Medical Insurance Payment Method

The patients with CKD with medical insurance were more likely to be in Profile 3. Various medical insurance schemes, to some degree, can alleviate the financial burden on patients with CKD. Medical insurance coverage can effectively reduce the proportion of personal medical expenditure (Xu et al. 2018), enabling patients to maintain a relatively stable quality of living and to reduce the risk of poverty caused by illness (Min et al. 2018). Therefore, the sense of security brought by medical insurance can enable patients to focus more on treatment rather than on worrying about the cost, effectively alleviating anxiety, depression and other psychological symptoms (Che et al. 2021). However, in China, there are a minority of individuals not enrolled in medical insurance, those requiring treatments or medications not covered by insurance, or those seeking care outside their insured region without proper authorization must bear their medical expenses out‐of‐pocket. For the self‐paying patients, the high cost of treatment was often a heavy financial burden, which can directly lead to a decline in their quality of life. The economic pressure not only exacerbated the physiological and psychological symptoms of patients but also hindered physical activity (Molsted et al. 2021), adversely affecting disease control and long‐term health.

Therefore, medical insurance with a high reimbursement rate and more financial help is suggested to be provided to the self‐paying and low‐income family (Guo et al. 2022). However, such support requires complex coordination of healthcare services from policymakers and different stakeholders, such as social charities and multi‐level commercial medical insurance, to jointly make the effort to increase the reimbursement rate. At present, what nephrology nurses can do is to identify patients with CKD who are economically burdened and refer these patients to social workers or some charities for assistance (He and Bian 2024; Liu and Chu 2024).

4.2.3. Daily Care

The results of the study showed that patients who did not need help from others for daily care had mild symptom distress and functional impairments and that patients who needed other caregivers had severe symptom distress and functional impairments. Daily care behaviours for patients with CKD included adherence to medication, dietary restrictions, regular exercise and abstaining from smoking and alcohol (Moattari et al. 2012). These behaviours decelerated the decline of kidney function, reduced proteinuria, controlled blood pressure and lowered C‐reactive protein levels to alleviate physical symptoms such as pain as well as psychological symptoms such as depression (Fraser and Blakeman 2016; Niu et al. 2021; Lee et al. 2016). However, patients who required help from others in daily care, particularly those requiring long‐term dialysis or transplantation, were often with severe symptoms and functional impairments and required the care from others (Al Muchtari et al. 2023; Vettoretti et al. 2020; Høy et al. 2007; Jaberi et al. 2022). They were unable to engage in physical and social activities due to increased physical symptoms, such as pain, which in result exacerbated psychological symptoms such as anxiety and depression (Palmer et al. 2024).

To address these issues, several key considerations arise. First, healthcare professionals could identify CKD patients with good daily care by themselves' behaviours and encourage them to share their experiences, especially on how to effectively adopt the health management behaviours in daily care, in order to encourage other patients to engage in their daily care. Second, healthcare professionals could form the patient clubs, where patients with CKD and their family members are encouraged to actively participate in various activities. Patient clubs may help improve patients' social function and reduce physiological and psychological symptoms. Finally, to successfully transfer healthcare related knowledge into patients' daily care behaviours, healthcare professionals could introduce behavioural facilitation models such as Fogg's Behaviour Model (Alrige et al. 2021), which focuses on motivation, ability and triggers. The motivation can be increased by linking daily care actions, like following a diet or taking medication, to personal health goals and rewarding progress. The ability can be enhanced through daily care training and support. The triggers, such as timely reminders or social support, can prompt patients to take necessary actions. By using this approach, healthcare providers can empower patients to adopt and sustain daily care behaviours for better health outcomes.

4.2.4. Diabetes

CKD patients with diabetes were more likely to be in Profile 1. In developed nations, diabetes has become the leading cause of end‐stage CKD, with approximately 50% of cases of end‐stage CKD closely linked to diabetes (Winocour 2018; Tuttle et al. 2014). This indicates that as the severity of kidney disease increases, the risk of concurrent diabetes also arises significantly (Gómez‐García et al. 2024). Patients with diabetic nephropathy had severe symptoms mainly because of metabolic disorders caused by high glycemic levels and increased glomerular filtration rate. These problems exacerbated the kidney's burden and the progression of CKD (Marassi and Fadini 2023; Tonneijck et al. 2017). Furthermore, diabetic nephropathy not only increased the risk of cardiovascular complications but also severely diminished patients' quality of life, imposing heavy psychological and physical burdens (Marassi and Fadini 2023; Tonelli et al. 2012) (Kim et al. 2022).

To address this challenge, glycemic control was of paramount importance (Copur et al. 2020). Research has shown that maintaining stable blood glucose levels not only improves physical health by reducing fatigue, improving energy levels and supporting physical activities (Ducat et al. 2014). Effective blood glucose management helps reduce psychological symptoms such as anxiety and depression, which can lead to better engagement in social activities and improved social interactions. Additionally, dietary factors played an irreplaceable role in preventing the occurrence and progression of CKD and diabetes (Sepandi et al. 2022). At present, the diet‐related health recommendation system (HRS) is gradually becoming a new hot spot in the field of dietary management. These HRS, by collecting and analysing personalised data from patients, can provide tailored recommendations from a wide range of options (Beij 2020). Most importantly, when these recommendations incorporate professionals' guidance from nutrition experts, they motivate patients to more actively follow healthy dietary habits (Gómez‐García et al. 2024; Mattei and Alfonso 2020). Research shows that HRS, aligned with patients' conditions and preferences, can offer up‐to‐date advice that adjusts as their health evolves (Cai, Yu, et al. 2022; Yera et al. 2023). In view of this, we strongly recommend that healthcare professionals develop diet‐related HRS typically for patients with CKD, as these systems could help control the level of blood glucose, slow the progression of CKD, and maintain patients' overall health.

4.2.5. Anaemia

The patients with CKD who were diagnosed with anaemia were more likely to be in Profile 1 and Profile 2. The primary cause of anaemia in patients with CKD is the relative deficiency of erythropoietin, a kind of hormone primarily produced by adult kidneys. Additionally, shortened red blood cell lifespan and functional iron deficiency also contribute to anaemia in CKD (Babitt and Lin 2012). As CKD progressed, the incidence of anaemia increased dramatically, from 8.4% in Stage 1 to 53.4% in Stage 5 (van Haalen et al. 2020). Anaemia can occur at any stage of the disease, causing physical symptoms like fatigue and pain and psychological symptoms like depression. As a result, anaemia can affect patients' physical and social function (van Haalen et al. 2020; Retat et al. 2024), because patients with these symptoms lack the energy required for physical and social activities (Farag et al. 2023). Therefore, healthcare professionals should remind patients of following medical advice regarding iron supplementation, should monitor haemoglobin levels, and should encourage patients to actively participate in physical activities to enhance physical endurance and alleviate the worsening of anaemia symptoms.

4.2.6. Disturbance of Phosphorus Metabolism

The patients with CKD experiencing disturbances in phosphorus metabolism were more likely to be in Profile 1. As a central issue in CKD‐Mineral and Bone Disorder (CKD‐MBD), the imbalance of phosphorus homeostasis stimulates systemic mineral metabolism abnormalities, bone changes, and vascular calcification (Biruete et al. 2023). These conditions significantly increased the risks of fractures, cardiovascular events and mortality, thereby intensifying symptom distress and functional impairments (Isakova et al. 2017). Studies have shown that serum phosphorus changes appeared early in CKD and worsen as kidney function further declines (Inker et al. 2019). Across different stages of CKD, the prevalence of hyperphosphatemia increased dramatically from 32.8% in Stage 3 to 83.6% in Stage 5 (Vikrant and Parashar 2016). This shift reflected that the kidney's inability to effectively excrete excess phosphorus, which in turn accelerates the decline in kidney function, a process that exacerbated symptoms and functional impairments (Kuro 2019). Currently, the primary clinical strategies for managing hyperphosphatemia in CKD involved haemodialysis, phosphate binders and dietary phosphorus restrictions (Feng et al. 2022). Among these, dietary control played an important role (Joshi et al. 2018), and the recommendations for dietary management are as aforementioned for glycemic control.

4.3. Limitation

First, the collection of data in only one hospital in Shanghai, China, may limit the generalizability of the findings. Second, the relatively limited sample size may have reduced the statistical power to detect significant differences in symptom distress and functional impairments across different CKD stages. In particular, while a three‐profile solution was identified through latent profile analysis, the small sample size may limit the robustness and external validity of this classification. Larger and more representative samples are needed in future studies to validate these findings. Third, PROMIS‐29 V2.1 was primarily used to evaluate generic symptom distress but did not include the CKD‐specific symptoms. Further exploration of CKD‐specific symptoms will be conducted in the future. Finally, due to the cross‐sectional design of the study, it was impossible to track and analyse the trajectories of symptoms and functional changes over time. To gain a deeper understanding of how symptoms and functions change, future research should employ a longitudinal study design. Additionally, it is recommended that future studies incorporate CKD‐specific assessment tools to complement the limitations of the PROMIS.

5. Conclusion

Via the PROMIS‐29 tool, three distinct latent profiles were identified among Chinese patients with chronic kidney disease (CKD) concerning symptoms and functions, including ‘moderate to severe symptom distress and moderate to severe functional impairments’, ‘low symptom distress and no functional impairments’ and ‘no symptom distress and no functional impairments’. Sociodemographic characteristics including ages, payment methods, daily care by myself and medical information such as diabetes, anaemia and disturbance of phosphorus metabolism had significant associations with the latent profile membership. Understanding these profiles and the factors predicting membership could provide valuable guidance for healthcare professionals in identifying the patients who are more likely to need additional support, and thereby more effective symptom management and functional improvement interventions can be tailored for patients with CKD.

Author Contributions

Study conception and design: Jingting Wang and Guoxiang Liu. Data collection: Guoxiang Liu, Qun Gu and Jinling Zhao. Data analysis and interpretation: Bei Yun, Xuanyi Bi, Qun Gu and Jingting Wang. Drafting of the article: Bei Yun. Critical revision of the article: Jingting Wang.

Funding

This study was funded by Soft Science Research Project of Shanghai (grant number: 25692108400).

Disclosure

The authors affirm that the statistical methods used were appropriate to the study design and data, and that the analyses were conducted and interpreted correctly. The authors take full responsibility for the integrity and accuracy of the statistical analysis.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.