Abstract
The purpose of this study was to determine the relationship between the serum levels of kappa (κ) and lambda (λ) total light chain (TLC), the κ/λ ratio, and the combined serum TLCκ and TLCλ (ΣTLC) levels in Chinese centenarians and all-cause mortality. The association between serum TLC and all-cause mortality was investigated using restricted cubic spline (RCS) analysis, Cox proportional hazards models, and Kaplan‒Meier curves. The study included 906 centenarians (18.8% male), 838 (92.5%) of whom died during a median follow-up of 30 months. The proportions of centenarians with abnormal TLCκ and TLCλ levels were 68.1% and 49.0%, respectively. RCS analysis indicated that the levels of TLCκ, TLCλ, and ΣTLC were associated with all-cause mortality (P < 0.05), whereas the κ/λ ratio was not (P > 0.05). Cox proportional hazards analysis demonstrated that the highest quartiles of TLCκ, TLCλ, and ΣTLC were associated with an increased risk of death, with hazard ratios of 1.434 (95% confidence interval [CI], 1.061–1.939; P = 0.019), 1.351 (95% CI, 1.013–1.802; P = 0.041), and 1.891 (95% CI, 1.347–2.654; P < 0.001), respectively. Kaplan–Meier analysis illustrated that centenarians with higher levels of TLCκ, TLCλ, and ΣTLC had significantly shorter median survival times (26 months versus 35 months, P < 0.001; 26 months versus 32 months, P = 0.003; and 26 months versus 36 months, P < 0.001, respectively). Our findings suggest that serum levels of TLCκ, TLCλ, and ΣTLC are significantly associated with all-cause mortality in centenarians.
Trial registration Not applicable.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12979-025-00520-0.
Keywords: Centenarians, Immunity, Aging, Longevity, Mortality risk, Kappa, Lambda
Introduction
Human aging is an exceedingly complex process characterized by the decline of physiological functions over time, leading to a deterioration in quality of life [1]. It is a universal biological phenomenon and an inevitable trend in the course of life. Within the context of global population aging, individuals experience the aging process differently based on genetic and environmental factors; some individuals are able to live exceptionally long and healthy lives. Centenarians represent a unique population because many maintain good functional status for a large portion of their lives, making them an ideal model for studying healthy aging. Some studies suggest that their status may be attributable to the robust immune function and defense mechanisms exhibited by centenarians [2]. The immune system has long been regarded as a hallmark of health [3] because it plays a crucial role in maintaining overall well-being.
Aging is characterized by progressive immune dysregulation, immunosenescence itself is a part of the aging process, and at the same time, it will accelerate the aging of other tissues and organs. The number of B cells may remain relatively stable in the elderly, but their functions have undergone significant changes. B cells and plasma cells secrete immunoglobulins, which are primarily composed of two heavy chains and two light chains linked by disulfide bonds. The light chains can be further classified into κ (kappa) and λ (lambda) light chains. Light chains that bind to heavy chains are referred to as bound light chains, whereas those that do not bind to heavy chains are called free light chains (FLC). The sum of bound light chains and FLC constitutes total light chains (TLC). The concentration of FLC has been widely used to monitor the activity of the adaptive immune system and can serve as an effective predictive indicator for the health and disease status of different patient populations, such as those affected by B cell- and plasma cell-related diseases [4, 5]. In non-diseased populations, elevated FLC has also been linked to mortality, suggesting systemic immune activation as a shared mechanism [6]. Unlike FLC, there are no studies about the relationship between TLC and all-cause mortality in non-disease population. However, TLC reflects the total immunoglobulin turnover, offering a comprehensive measure of B-cell and plasma cell activity. This is particularly relevant in aging, where dysregulated immune responses and chronic inflammation may drive morbidity. Meanwhile, the detection of TLC is cheaper than that of FLC, and many primary hospitals have not yet carried out the detection of FLC. This study specifically focuses on the relationships between the levels of TLCκ and TLCλ in centenarians and all-cause mortality. Through this study, we aim to provide novel insights into the role of the immune system in the process of extreme aging and offer a scientific basis for future intervention strategies targeting the health of the elderly population.
Methods
Study design and population
This study used data from the China Hainan Centenarian Cohort Study, a nationally representative prospective cohort study of centenarians conducted in Hainan Province between July 2014 and December 2016. A total of 1,002 centenarians participated in this study, and the baseline data have been reported in previous studies [7, 8]. To ensure data accuracy, death dates and causes were rigorously validated through a tripartite verification process: Official Records: Cross-referenced with the National Cause of Death Registration and Reporting Information System (China CDC); Administrative Confirmation: Validated by local civil affairs authorities; Family Verification: Confirmed via structured telephone interviews with next of kin. Additionally, survival status for centenarians was monitored monthly by the Hainan Provincial Civil Affairs Bureau, which administers pension disbursements for individuals aged ≥ 80 years, ensuring real-time updates on participant vitality. The study was approved by the Ethics Committee of the Chinese People’s Liberation Army (PLA) General Hospital (No. 301HNLL-2016-01), and was conducted in accordance with the Declaration of Helsinki and its subsequent revisions. All the participants provided signed informed content and agreed to attend this survey. This study was designed and reported in accordance with the STROBE guidelines for observational studies.
Covariates
The study cohort of centenarian participants received a comprehensive health assessment at Hainan Hospital of PLA General Hospital, where they participated in structured interviews conducted by a multidisciplinary clinical team. Systematic data collection included demographic parameters (age, gender, ethnic background, marital status, and educational attainment), lifestyle factors (tobacco use and alcohol consumption patterns), and medical history documentation (diabetes mellitus [DM], hypertension, and coronary heart disease [CHD]) that were obtained via structured interviews with standardized questionnaires. Hypertension, diabetes mellitus, and CHD were defined as self-reported diagnoses or use of disease-specific medications. Hypertension was diagnosed based on systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg. Diabetes mellitus was defined according to the 1999 World Health Organization (WHO) diagnostic criteria for diabetes [9]. Anthropometric measurements of weight and height were obtained using standardized protocols, with body mass index derived from the weight-to-height squared ratio (kg/m2). The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation [10].
Blood samples were drawn from participants by an experienced nurse and sent to the Laboratory Department of Hainan Hospital of the PLA General Hospital. Serum TLCκ and TLCλ were quantified using a fully automated protein analyzer (BNII; Siemens AG, Munich, Germany) with the light-scattering turbidimetric immunoassay method. The normal concentrations of TLCκ and TLCλ, as well as the TLC ratio, are defined, in accordance with the manufacturer’s recommendations, as follows: the normal κ light chain concentration range is 1.7–3.7 g/L, the range for λ light chain is 0.9–2.1 g/L, and the κ/λ ratio is 1.35–2.65. ΣTLC means the combined serum TLCκ and TLCλ.
Statistical analysis
Data were tested for normality and homogeneity of variance before statistical analysis was performed. Normally distributed data were described as mean ± standard deviation. The independent samples t-test or analysis of variance was used to analyze differences between groups. Asymmetrically distributed data were described as the median and interquartile range (IQR) and the Mann–Whitney U test or Kruskal–Wallis H test was used to analyze differences between groups. Categorical variables were described as numbers with percentages and compared with the χ2 test. Spearman correlation was used to analyze the correlations among various variables. Following conventional criteria, Spearman correlation coefficients were categorized as follows: 0.00-0.39 (weak), 0.40–0.69 (moderate), 0.70–1.00 (strong) [11].
The four TLC-based predictors, TLCκ, TLCλ, κ/λ and ΣTLC, were examined to find the relationship with the all-cause mortality using both unadjusted and multivariate-adjusted models. Restricted cubic spline (RCS) analysis is widely used to study nonlinear associations [12]. RCS analyses were performed to assess the associations between all-cause mortality and serum TLCκ, TLCλ, κ/λ, and ΣTLC levels as continuous variables in both unadjusted and multivariate-adjusted models. The RCS were built using the R package rms, with 4 knots placed. The placement of knots followed the default settings that are the 5th, 35th, 65th, and 95th percentiles of the studied variable. The quartile breaks for the TLC-based predictors are as follows: 3.54, 4.13, and 4.78 g/L for TLCκ; 1.78, 2.09, and 2.49 g/L for TLCλ; 1.75, 1.97, and 2.19 for κ/λ; 5.40, 6.24, and 7.20 g/L for ΣTLC. Prior to performing Cox regression analyses, we conducted tests for the proportional hazards assumption for serum TLCκ, TLCλ, κ/λ, and ΣTLC. Both univariate and multivariate Cox regression analyses were conducted in this study: the former to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for mortality, and the latter to adjust for confounding factors. Based on data availability, previous studies, and potential links to aging and immune function, we adjusted the following variables. Multivariate model was adjusted for age, gender, ethnicity, marital status, body mass index, education, smoking status, alcohol status, and status for diabetes mellitus, hypertension, eGFR, CHD, serum immunoglobulin (Ig) M, IgG, IgA, IgE, C3, and C4. When taking TLCκ or TLCλ as the predictor, TLCλ or TLCκ was also included in the model. The distribution of time to death is shown in Kaplan–Meier survival curves and compared using the log-rank test. The robustness of the observed association was assessed using E-value analysis, which quantifies the minimum strength of association that unmeasured confounders (e.g., healthcare access, socioeconomic status) would need to have with both the exposure and outcome to nullify the observed relationship [13]. In addition, subgroup analyses were performed to estimate whether the subgroups exhibit different relationships, and the P-values for interaction were evaluated through likelihood ratio tests. Differences for which P < 0.05 were considered statistically significant. Data processing was performed in R language software (version 4.3.3).
Results
Characteristics of the study population at baseline and during follow-up
We excluded patients with monoclonal gammopathy and other hematological disorders, as well as centenarians with chronic infectious diseases and incomplete data (N = 96) (Fig. 1). Table 1 shows characteristics of the study population based on the sum of TLCκ and TLCλ (ΣTLC) quartiles at baseline. The median age was 102 years (IQR, 101–104), and 18.8% were male. Nearly three-quarters (74.0%) had hypertension, and smaller proportions had diabetes mellitus (9.1%) or CHD (4.4%). The serum TLCκ levels of the majority of participants (68.1%) were above the normal range, and the remaining participants had serum TLCκ levels within the normal range (31.9%). Approximately half of the participants had serum TLCλ within the normal range (51%), and the other half had serum TLCλ above the normal range (49%).
Fig. 1.
Flow chart of subject selection
Table 1.
Characteristics of the study population based on Σ TLC quartiles at baseline
| Variable | Overall | Q1[2.74,5.4) | Q2[5.4,6.24) | Q3[6.24,7.2) | Q4[7.2,12.3) | P |
|---|---|---|---|---|---|---|
| N | 906 | 227 | 223 | 229 | 227 | |
| Age, years | 102.00 (101.00, 104.00) | 102.00 (101.00, 104.00) | 102.00 (101.00, 104.00) | 102.00 (101.00, 104.00) | 102.00 (101.00, 104.00) | 0.593 |
| Male, % | 170 (18.80) | 31 (13.70) | 41 (18.40) | 52 (22.70) | 46 (20.30) | 0.087 |
| Follow-up time, months | 30.10 (14.80, 53.22) | 35.60 (16.05, 60.95) | 31.20 (17.75, 55.75) | 30.20 (12.70, 51.90) | 26.40 (14.20, 42.20) | 0.006 |
| Death, % | 838 (92.50) | 202 (89.00) | 205 (91.90) | 209 (91.30) | 222 (97.80) | 0.001 |
| Ethnicity | < 0.001 | |||||
| Han, % | 804 (88.70) | 215 (94.70) | 197 (88.3) | 206 (90.0) | 186 (81.9) | |
| Other, % | 102 (11.30) | 12 (5.30) | 26 (11.70) | 23 (10.00) | 41 (18.10) | |
| Marital status | 0.976 | |||||
| Married, % | 96 (10.60) | 25(11.00) | 22(9.90) | 24(10.50) | 25(11.00) | |
| Separation/Divorce/Widowhood, % | 810 (89.40) | 202 (89.00) | 201 (90.10) | 205 (89.50) | 202 (89.00) | |
| Education | 0.233 | |||||
| No education, % | 824 (90.90) | 208 (91.60) | 195 (87.40) | 208 (90.80) | 213 (93.80) | |
| Elementary school, % | 63 (7.00) | 17 (7.50) | 21 (9.40) | 15 (6.60) | 10 (4.40) | |
| Junior high school and above, % | 19 (2.10) | 2 (0.90) | 7 (3.10) | 6 (2.60) | 4 (1.80) | |
| Smoke | 0.185 | |||||
| Never, % | 806 (89.00) | 208 (91.60) | 197 (88.30) | 198 (86.50) | 203 (89.40) | |
| Past, % | 68 (7.50) | 15 (6.60) | 13 (5.80) | 22 (9.60) | 18 (7.90) | |
| Now, % | 32 (3.50) | 4 (1.80) | 13 (5.80) | 9 (3.90) | 6 (2.60) | |
| Drinking | < 0.001 | |||||
| Never, % | 746 (82.30) | 199 (87.70) | 190 (85.20) | 191 (83.40) | 166 (73.10) | |
| Past, % | 71 (7.80) | 11 (4.80) | 19 (8.50) | 23 (10.00) | 18 (7.90) | |
| Now, % | 89 (9.80) | 17 (7.50) | 14 (6.30) | 15 (6.60) | 43 (18.90) | |
| Hypertension, % | 670 (74.00) | 158 (69.60) | 173 (77.60) | 166 (72.50) | 173 (76.20) | 0.203 |
| Diabetes mellitus, % | 82 (9.10) | 27 (11.90) | 16 (7.20) | 22 (9.60) | 17 (7.50) | 0.268 |
| Coronary heart disease, % | 40 (4.40) | 8 (3.50) | 14 (6.30) | 8 (3.50) | 10 (4.40) | 0.437 |
| Body mass index, kg/m2 | 18.15 (16.22, 20.08) | 17.90 (16.23, 19.99) | 18.38 (16.44, 20.02) | 18.44 (16.63, 20.39) | 17.65 (15.51, 19.66) | 0.037 |
| eGFR, ml/min/1.73m2 | 55.64 (43.48, 68.35) | 57.46 (43.38, 68.38) | 54.11 (42.85, 68.97) | 54.83 (44.43, 65.44) | 57.20 (43.63, 69.09) | 0.706 |
| eGFR Categories | 0.57 | |||||
| eGFR < 60 | 545 (60.2) | 132 (58.1) | 135 (60.5) | 146 (63.8) | 132 (58.1) | |
| eGFR ≥ 60 | 361 (39.8) | 95 (41.9) | 88 (39.5) | 83 (36.2) | 95 (41.9) | |
| IgM, g/L | 1.02 (0.71, 1.40) | 0.92 (0.66, 1.33) | 1.04 (0.72, 1.33) | 0.97 (0.72, 1.38) | 1.13 (0.76, 1.49) | 0.008 |
| IgG, g/L | 15.70 (13.60, 18.10) | 12.90 (11.65, 14.05) | 14.70 (13.60, 16.10) | 16.90 (15.20, 18.10) | 19.70 (17.60, 22.00) | < 0.001 |
| IgA, g/L | 3.33 (2.53, 4.25) | 2.45 (1.98, 2.96) | 3.22 (2.51, 3.92) | 3.64 (2.89, 4.44) | 4.30 (3.52, 5.28) | < 0.001 |
| IgE, IU/mL | 273.00 (81.50, 814.75) | 136.00 (46.00, 437.00) | 288.00 (83.50, 746.00) | 310.00 (106.00, 1090.00) | 379.00 (144.00, 1065.00) | < 0.001 |
| Serum C3, g/L | 0.98 (0.86, 1.11) | 0.99 (0.84, 1.12) | 0.94 (0.84, 1.06) | 0.98 (0.89, 1.11) | 1.00 (0.87, 1.17) | 0.006 |
| Serum C4, g/L | 0.23 (0.19, 0.28) | 0.23 (0.19, 0.28) | 0.23 (0.19, 0.29) | 0.23 (0.19, 0.27) | 0.23 (0.19, 0.28) | 0.855 |
| TLCκ, g/L | 4.13 (3.54, 4.78) | 3.17 (2.94, 3.39) | 3.87 (3.67, 4.08) | 4.40 (4.21, 4.60) | 5.31 (4.94, 5.81) | < 0.001 |
| TLCκ Categories | < 0.001 | |||||
| Normal group [1.7–3.7 g/L] | 255 (32.10) | 224 (98.70) | 65 (28.30) | 2 (0.90) | 0 (0.00) | |
| High group(> 3.7 g/L) | 615 (67.9) | 3 (1.3) | 158 (70.9) | 227 (99.1) | 227 (100.0) | |
| TLCλ, g/L | 2.09 (1.78, 2.49) | 1.66 (1.47, 1.81) | 1.95 (1.78, 2.09) | 2.23 (2.06, 2.40) | 2.69 (2.51, 3.04) | < 0.001 |
| TLCλ Categories | < 0.001 | |||||
| Normal group [0.9–2.1 g/L] | 466 (51.40) | 221 (97.40) | 171 (76.7) | 67 (229.3) | 7 (3.10) | |
| High group(> 2.1 g/L) | 440 (48.6) | 6 (2.6) | 52 (23.3) | 162 (70.7) | 220 (96.9) | |
| κ/λ | 1.97 (1.75, 2.19) | 1.92 (1.71, 2.14) | 2.02 (1.79, 2.25) | 1.96 (1.78, 2.21) | 1.97 (1.75, 2.17) | 0.044 |
| Σ TLC | 6.24 (5.40, 7.20) | 4.83 (4.50, 5.12) | 5.81 (5.61, 6.02) | 6.65 (6.45, 6.89) | 7.94 (7.52, 8.61) | < 0.001 |
Correlation analyses of serum TLCκ, TLCλ, κ/λ, and ΣTLC with other variables
Except for moderately strong correlations between education and gender (r = 0.402, P < 0.001), between smoking and gender (r = 0.462, P < 0.001), and between C3 and C4 (r = 0.403, P < 0.001), all the Spearman correlation coefficients among age, gender, ethnicity, marital status, education, smoking status, drinking status, hypertension, DM, CHD, BMI, eGFR, serum immunoglobulin (Ig) M, IgG, IgA, IgE, C3, and C4 were < 0.4, suggesting a weak association among these indicators (Fig. 2 and Supplementary Table 1). There was a strong association between TLCκ and IgG (r = 0.769, P < 0.001), a moderate association with IgA (r = 0.539, P < 0.001), and a weak association with IgM, IgE, C3, and C4. Similar to TLCκ, there was a strong association between TLCλ and IgG (r = 0.711, P < 0.001), a moderate association with IgA (r = 0.558, P < 0.001), and a weak association with IgM, IgE, C3 and C4. There is a moderate correlation between TLCκ and TLCλ (r = 0.695, P < 0.001). κ/λ has a moderately strong association with TLCλ (r = -0.430, P < 0.001), but only a weak association with TLCκ, IgM, IgG, IgA, IgE, C3 and C4, suggesting that this indicator can only reflect TLCλ. ΣTLC has a strong correlation with TLCκ (r = 0.960, P < 0.001), TLCλ (r = 0.861, P < 0.001), and IgG (r = 0.807, P < 0.001), and a moderate correlation with IgA (r = 0.587, P < 0.001). It is indicated that this indicator not only can reflect the levels of TLCκ and TLCλ, but also the ability to reflect the levels of IgG and IgA.
Fig. 2.
Correlation matrix of demographic, lifestyle, clinical, and immunological variables. Values represent correlation coefficients, the color of the font represents relevance. Negative values denote inverse correlations, while positive values indicate direct relationships. Following conventional criteria, Spearman correlation coefficients were categorized as follows: 0.00-0.39 (weak), 0.40–0.59 (moderate), 0.60–1.00 (strong).The r-values and p-values for each pairwise comparison of variables can be found in Supplementary Table 1 (Supplementary Material)
RCS analyses of serum TLCκ, TLCλ, κ/λ, and ΣTLC
Figure 3 presents the RCS analysis results for serum TLCκ, TLCλ, and ΣTLC. Serum TLCκ (Fig. 3A–B), TLCλ (Fig. 3C–D), and ΣTLC (Fig. 2G–H) are significantly associated with all-cause mortality in centenarians (all P < 0.05), whereas the κ/λ ratio is not associated with centenarian mortality (P > 0.05) (Fig. 2E–F). The relationships of all-cause mortality with TLCκ and TLCλ levels were significantly nonlinear in both the unadjusted and multivariable-adjusted RCS models (all Pnonlinear < 0.05). By contrast, ΣTLC is linearly associated with all-cause mortality risk in centenarians (univariate analysis: Pnonlinear = 0.073, and multivariate-adjusted analysis: Pnonlinear = 0.053; Fig. 2G–H). The association of TLCκ and TLCλ levels with the all-cause mortality rate of centenarians presented smooth “S” shapes, while the relationship of κ/λ and ΣTLC with all-cause mortality presented nearly linear shapes without a clear point where the risk changed direction.
Fig. 3.
Restricted cubic splines (RCS) of all-cause mortality with TLCκ, TLCλ, κ/λ, and ΣTLC. The solid red line indicated the hazard ratio, and the shaded areas represent the 95% CI. The horizontal coordinates indicated TLCκ, TLCλ, κ/λ, and ΣTLC levels, the left vertical coordinates indicated the hazard ratios for all-cause mortality, and the right vertical coordinates indicated the percentages of centenarians with serum TLCκ, TLCλ, κ/λ and ΣTLC levels for the histogram. Q1, Q2, Q3, Q4 are quartile spacing. Multivariate-adjusted analyses adjusted for age, gender, ethnicity, marital status, body mass index, education, smoking status, alcohol status, diabetes mellitus, hypertension, eGFR, coronary artery disease, serum C3, serum C4, IgM, IgG, IgA, and IgE. When taking TLCκ or TLCλ as the predictor, TLCλ or TLCκ was also included in the model. The reference cut-off values of serum κ, λ, and κ/λ were median values, and the reference cut-off values of serum ΣTLC was 75th percentile. The relationship of TLCκ and TLCλ with all-cause mortality was significantly no-linear in both unadjusted (A and C) and multivariable-adjusted (B and D) RCS models, with the serum TLCκ and TLCλ were significantly associated with all-cause mortality in centenarians (all the P < 0.05). The relationship of κ/λ and ΣTLC with all-cause mortality was significantly linear in both unadjusted (E and G) and multivariable-adjusted (F and H) RCS models. Serum κ/λ ratio was not associated with centenarian mortality (P > 0.05), and serum ΣTLC was significantly associated with all-cause mortality in centenarians (P < 0.05). The RCS curves of A, B, C and D present a smooth S-shaped form, while E, F, G and H are remains relatively stable without a clear point where the risk changes direction
Univariate and multivariate Cox analyses and Kaplan–Meier curves based on serum TLCκ, TLCλ, κ/λ, and ΣTLC levels
Table 2 provides a detailed illustration of the relationships between serum TLCκ, TLCλ, κ/λ, and ΣTLC in centenarians and the risk of all-cause mortality. In both univariate and multivariate analyses, the higher the levels of serum TLCκ, TLCλ, and ΣTLC, the higher the mortality rate of centenarians, whereas the κ/λ ratio does not show this trend. Among these measures, ΣTLC was associated with the highest mortality risk in both univariate and multivariate analyses. We used the group with the lowest IQR (Q1) as the reference. For serum TLCκ, the mortality risks of the Q3 and Q4 groups were higher than that of Q1, and the highest risk was for Q4 (multivariate-adjusted analysis: HR, 1.434; 95% CI, 1.061–1.939; P = 0.019). For serum TLCλ, the mortality risks of the Q4 group was higher than that of Q1(multivariate-adjusted analysis: HR, 1.351; 95% CI, 1.013–1.802; P = 0.041). For serum ΣTLC, the mortality risks of the Q4 group was higher than that of Q1 (multivariate-adjusted analysis: HR, 1.891; 95% CI, 1.347–2.654; P < 0.001). For serum TLCκ/λ, there was no statistically significant association with all-cause mortality risk in the centenarian population. The E-values for the TLCκ (Q1-Q2 versus Q3-Q4), TLCλ (Q1-Q2 versus Q3-Q4), κ/λ (Q1-Q2 versus Q3-Q4), and ΣTLC (Q1-Q3 versus Q4) were 1.854 (1.481, Infinity), 1.571 (1.160, Infinity), 1.100 (Infinity, 1.000), and 1.934 (1.476, Infinity), respectively. A higher E-value indicates greater robustness of the association. The E-value for the ΣTLC was the biggest, indicating that the association between serum ΣTLC and mortality was the most robust.
Table 2.
Univariate and multivariate Cox proportional hazards analyses of the associations of TLCκ, TLCλ, κ/λ, and ΣTLC with all-cause mortality in centenarians
| Number | Univariate analysis | Multivariate-adjusted analysis | |||
|---|---|---|---|---|---|
| HR | P | HR | P | ||
| TLCκ per 1 g/L | 906 | 1.127(1.055,1.205) | < 0.001 | 1.174(1.024,1.346) | 0.021 |
| Group by Interquartile Value | |||||
| Q1 [1.72,3.54) | 226 | 1 | 1 | ||
| Q2 [3.54,4.13) | 225 | 0.952(0.783,1.157) | 0.619 | 0.939(0.759,1.161) | 0.560 |
| Q3 [4.13,4.78) | 223 | 1.264(1.042,1.532) | 0.017 | 1.329(1.051,1.679) | 0.017 |
| Q4 [4.78,7.64] | 232 | 1.358(1.122,1.644) | 0.002 | 1.434(1.061,1.939) | 0.019 |
| P for trend | < 0.001 | 0.003 | |||
| Group by P50 Value | |||||
| Q1-Q2 [1.72,4.13) | 451 | 1 | 1 | ||
| Q3-Q4 [4.13,7.64] | 455 | 1.343(1.172,1.539) | < 0.001 | 1.411(1.175,1.693) | < 0.001 |
| E-values | 1.754(1.476, Inf) | 1.854(1.481, Inf) | |||
| TLCλ per 1 g/L | 906 | 1.174(1.043,1.322) | 0.008 | 1.088(0.882,1.341) | 0.432 |
| Group by Interquartile Value | |||||
| Q1 [0.93,1.78) | 227 | 1 | 1 | ||
| Q2 [1.78,2.09) | 222 | 0.964(0.794,1.171) | 0.713 | 1.007(0.817,1.241) | 0.946 |
| Q3 [2.09,2.49) | 228 | 1.147(0.947,1.389) | 0.162 | 1.197(0.954,1.502) | 0.121 |
| Q4 [2.49,4.71] | 229 | 1.337(1.105,1.618) | 0.003 | 1.351(1.013,1.802) | 0.041 |
| P for trend | < 0.001 | 0.022 | |||
| Group by P50 Value | |||||
| Q1-Q2 [0.93,2.09) | 449 | 1 | 1 | ||
| Q3-Q4 [2.09,4.71] | 457 | 1.258(1.098,1.441) | < 0.001 | 1.227(1.028,1.465) | 0.023 |
| E-values | 1.622(1.334, Inf) | 1.571(1.160, Inf) | |||
| κ/λ per 1 unit | 906 | 1.067(0.894,1.272) | 0.473 | 1.050(0.873,1.263) | 0.607 |
| Group by Interquartile Value | |||||
| Q1 [1,1.75) | 227 | 1 | 1 | ||
| Q2 [1.75,1.97) | 226 | 1.067(0.881,1.293) | 0.505 | 1.103(0.907,1.340) | 0.325 |
| Q3 [1.97,2.19) | 226 | 1.008(0.831,1.223) | 0.932 | 1.023(0.841,1.244) | 0.823 |
| Q4 [2.19,3.98] | 227 | 1.087(0.898,1.316) | 0.392 | 1.082(0.886,1.320) | 0.441 |
| P for trend | 0.530 | 0.623 | |||
| Group by P50 Value | |||||
| Q1-Q2 [1,1.97) | 453 | 1 | 1 | ||
| Q3-Q4 [1.97,3.98] | 453 | 1.014(0.885,1.161) | 0.846 | 1.000(0.870,1.149) | 0.999 |
| E-values | 1.109(1.000, Inf) | 1.100(Inf, 1.000) | |||
| Σ TLC per 1 g/L | 906 | 1.084(1.036,1.135) | < 0.001 | 1.146(1.027,1.279) | 0.015 |
| Group by Interquartile Value | |||||
| Q1 [2.74,5.4) | 227 | 1 | 1 | ||
| Q2 [5.4,6.24) | 223 | 1.099(0.905,1.335) | 0.341 | 1.172(0.943,1.455) | 0.152 |
| Q3 [6.24,7.2) | 229 | 1.119(0.922,1.358) | 0.255 | 1.278(0.996,1.640) | 0.054 |
| Q4 [7.2,12.3] | 227 | 1.516(1.251,1.837) | < 0.001 | 1.891(1.347,2.654) | < 0.001 |
| P for trend | < 0.001 | 0.001 | |||
| Group by P75 Value | |||||
| Q1-Q3 [2.74,7.2) | 679 | 1 | 1 | ||
| Q4 [7.2,12.3] | 227 | 1.416(1.213,1.652) | < 0.001 | 1.468(1.172,1.839) | < 0.001 |
| E-values | 1.861(1.548, Inf) | 1.934(1.476, Inf) | |||
Multivariate-adjusted analysis adjusted for age, gender, ethnicity, marital status, body mass index, education, smoking status, alcohol status, diabetes mellitus, hypertension, coronary heart disease, estimated glomerular filtration rate (eGFR), serum IgM, IgG, IgA, IgE, C3, C4. When taking TLCκ or TLCλ as the predictor, TLCλ or TLCκ was also included in the model.
Figure 4 presents the Kaplan–Meier curves and P-values illustrating the differences in mortality rates based on the quartiles of serum TLCκ, TLCλ, κ/λ, and ΣTLC. The Kaplan–Meier curves demonstrate a significant correlation between higher levels of serum TLCκ, TLCλ, and ΣTLC and shorter survival times. For serum TLCκ, TLCλ, and ΣTLC levels, the median survival time of centenarians in the Q4 group was significantly shorter than that of individuals in the Q1 group (26 months versus 35 months, log-rank P < 0.001 for TLCκ; 26 months versus 32 months, log-rank P = 0.003 for TLCλ; and 26 months versus 36 months, log-rank P < 0.001 for ΣTLC). For serum TLCκ and TLCλ, the median survival time of centenarians in the Q3– Q4 group was significantly shorter than that of individuals in the Q1–Q2 group (27 months versus 35 months, log-rank P < 0.001; and 28 months versus 32 months, log-rank P = 0.003, respectively). For serum ΣTLC, the median survival time of centenarians in the Q4 group was significantly shorter than that of individuals in the Q1– Q3 group (26 months versus 32 months, log-rank P < 0.001).
Fig. 4.
Kaplan-Meier survival curves and log-rank tests for the associations of all-cause mortality with serum TLCκ, TLCλ, κ/λ, and ΣTLC levels. Kaplan-Meier survival curves revealed significant associations of higher TLCκ, TLCλ and ΣTLC levels with reduced survival time. The median survival time was significantly shorter in centenarians with the highest quartile of TLCκ than in those with the lowest quartile of TLCκ group (26 months vs. 35 months, P < 0.001) (A). The median survival time was significantly shorter for participants in the Q3-Q4 group than Q1-Q2 group on the serum κ group (27 months versus 35 months, P < 0.001) (B).The median survival time was significantly shorter in centenarians with the highest quartile of TLCλ than in those with the lowest quartile of TLCλ group (26 months vs. 32 months, P = 0.003) (C). The median survival time was significantly shorter for participants in the Q3-Q4 group than Q1-Q2 group on the serum TLCλ group (28 months versus 32 months, P < 0.001) (D). There was no significant difference in median survival time between different groups of serum κ/λ (all P > 0.05) (E, F).The median survival time was significantly shorter in centenarians with the highest quartile of ΣTLC than in those with the lowest quartile of ΣTLC group (26 months vs. 36 months, P < 0.001) (G). The median survival time was significantly shorter for participants in the Q4 group than Q1-Q3 group on the serum ΣTLC group (26 months versus 32 months, P < 0.001) (H)
Subgroup analyses of potential variables that could influence the association between TLCκ, TLCλ, κ/λ, and ΣTLC and all-cause mortality
Figure 5 provides forest plot analysis of adjusted HRs with 95% CIs confirmed the association between ΣTLC and mortality across predefined subgroups (Fig. 5). Each 1-unit (g/L) increase in ΣTLC was significantly associated with higher mortality risk (adjusted HR = 1.146, 95% CI: 1.027–1.279; P = 0.015). No significant interaction effects were observed across subgroups (all P for interaction > 0.05), indicating that the association between ΣTLC and mortality was consistent across the examined strata. Similarly, in the subgroup analyses of TLCκ, TLCλ and κ/λ, no significant interaction effects were observed among the subgroups (Supplementary Fig. 1–3).
Fig. 5.
Subgroup analysis of the association between ΣTLC and all-cause mortality. The forest plot presents hazard ratios with 95% confidence intervals across subgroups stratified by age, gender, ethnicity, marital status, hypertension, diabetes mellitus (DM), coronary heart disease (CHD), body mass index (BMI), estimated glomerular filtration rate (eGFR), serum IgM, IgG, IgA, IgE, C3, C4. P values for interaction between subgroups are provided. Each 1-unit (g/L) increase in ΣTLC was associated with a significantly increased risk of mortality in centenarians, with an adjusted HR of 1.146 (95% CI: 1.027–1.279, P = 0.005). Multivariate adjusted analysis was adjusted for age, gender, ethnicity, marital status, body mass index, education, smoking status, alcohol status, diabetes mellitus, hypertension, coronary artery disease, eGFR, serum IgM, IgG, IgA, IgE, C3, and C4. No significant interaction effects were observed among the subgroups, suggesting that the elevated risk associated with higher ΣTLC levels was consistent across different subgroup characteristics (all P for interaction > 0.05)
Discussion
This study is the first large-scale, long-term follow-up study on the relationship between serum TLC and all-cause mortality risk in centenarians. Our study not only found that higher serum TLCκ and TLCλ concentrations were associated with all-cause mortality risk in centenarians, but also explored the relationship of serum κ/λ and ΣTLC levels with all-cause mortality risk in centenarians. The κ/λ ratio was not associated with all-cause mortality rate in centenarians, whereas ΣTLC was highly correlated with centenarian all-cause mortality rate. After adjusting for population characteristics (e.g., age and gender), comorbidities (i.e., hypertension, diabetes, and CHD), eGFR, and other immune indicators in the multivariate Cox regression analysis, the all-cause mortality risks of centenarians in the highest quartiles of serum TLCκ, TLCλ, and ΣTLC levels were 1.434, 1.351, and 1.891 times higher, respectively, than those of centenarians in the lowest quartiles. These findings indicate that serum TLCκ, TLCλ, and ΣTLC can predict the overall mortality risk of centenarians and can serve as independent predictors of death.
Previous studies have shown that abnormally elevated FLC may serve as a useful diagnostic tool for certain blood diseases, such as multiple myeloma [14], diffuse large B cell lymphoma [15], primary amyloidosis [16], monoclonal gammopathy of undetermined significance [17]. Elevated FLC levels have also been considered significant in various non-hematologic diseases, such as diabetes [18] and chronic kidney disease [19], and inflammatory conditions, such as asthma [20] and systemic lupus erythematosus [21], suggesting that they may be biomarkers of B lymphocyte activation. Additionally, research on the relationship between serum FLC (κ and λ) and survival rate as well as mortality rate has mainly focused on populations with diseases. For example, Jin et al. found that serum FLC levels negatively correlated with survival rate in patients with light chain multiple myeloma [22]. Assi et al. also demonstrated that higher concentrations of FLCλ and FLCκ were associated with an increased risk of death in patients with chronic kidney disease [23]. Maurer et al. found that elevated serum FLC was an independent adverse prognostic factor for overall survival in diffuse large B cell lymphoma, and warrants further evaluation as a biomarker [15]. We studied light chain levels in centenarians and found nonlinear relationships between elevated serum TLCκ or TLCλ (not FLC) levels and risk of death in centenarians. Even after adjusting for multiple indicators, changes in these levels were still significantly associated with all-cause mortality risk.
Concurrently, we explored the relationship of the serum κ/λ ratio and ΣTLC with the all-cause mortality rate in centenarians. We found that the κ/λ ratio was not associated with all-cause mortality rate in centenarians, whereas ΣTLC strongly correlated with all-cause mortality. Under normal circumstances, the ratio of κ and λ light chains produced by B cells remains relatively stable. Abnormalities in the FLCκ/λ ratio are often used for the diagnosis of blood disorders, such as multiple myeloma and other plasma cell diseases. As a component of TLC, changes in the serum levels of FLC also affect the levels of TLC. Dispenzier et al. studied 15,859 residents of Olmsted County, Minnesota, who were 50 years old or older (after excluding persons with a known plasma cell disorder). In the general population without plasma cell disorders, an increase in non-clonal ΣFLC (the sum of serum FLCκ and FLCλ) was an important predictor of poorer overall survival [24]. However, the mechanism by which ΣFLC predicts survival rates in this large population remains undetermined. Additionally, Anandram et al. found that in a follow-up of 4.5 years, among 527 patients without monoclonal diseases, higher combined concentrations of serum FLCκ and FLCλ were an independent risk factor for death and could also identify patients at high risk of adverse outcomes in referral and hospital populations [25]. Similarly, the findings of our study highlight the importance of TLC in healthy aging and longevity, especially in centenarians, and identify potentially important biomarkers for assessing the health status of centenarians and predicting the risk of death. Although previous studies have mainly focused on serum FLC as biomarkers of B cell activation, our study expands this understanding by demonstrating that TLC, including both bound and unbound light chains, provide a comprehensive and accurate reflection of the dynamics of immunoglobulins in the body. This broader perspective is crucial for understanding the complex interplay between immune function and aging.
At present, in our study, the reason for the increase in the risk of all-cause mortality in centenarians associated with serum TLCκ, TLCλ, and ΣTLC levels remains unclear, although it may be related to the activation of the immune system in centenarians. Evidence suggests that an increase in immunoglobulin production is expected to also increase the synthesis of FLC [26]. The serum concentration of immunoglobulins and light chains is determined by the concentration of intact immunoglobulin molecules (IgG, IgA, IgM and IgE). In our previous study, we found that the majority of centenarians had normal serum IgM levels (93.3%), whereas small proportions had levels higher or lower than normal (3.0% and 3.8%, respectively). More than half of the participants had normal serum IgG levels (51.8%), and the remaining participants had serum IgG levels higher than normal (48.2%). The majority of participants had normal serum IgA levels (69.3%), and the remaining participants had serum IgA levels higher than normal (30.7%) [8]. Approximately one-quarter of the centenarians had normal serum IgE levels (28.3%), and the remaining three-quarters of participants had serum IgE levels higher than normal (71.7%).Therefore, in our study, the increase in serum TLCκ and TLCλ may be partly related to the higher levels of immunoglobulins in centenarians. However, it is worth noting that after adjusting for factors such as immunoglobulins in the multivariate analysis, the levels of serum TLCκ, TLCλ, and ΣTLC were still associated with mortality rate of centenarians, indicating that the relationships between serum TLCκ, TLCλ, and ΣTLC levels and all-cause mortality in centenarians are independent of immunoglobulin levels. In addition, the most common cause of increased FLC may be a reduction in clearance rate owing to kidney damage [27]. To further explore whether the relationship of serum TLCκ and TLCλ with all-cause mortality in centenarians is related to kidney function, we investigated this issue in the multivariate analysis.After adjusting for eGFR, the levels of serum TLCκ, TLCλ, and ΣTLC were still associated with the mortality rate of centenarians, indicating that the relationships between serum TLCκ, TLCλ, and ΣTLC and all-cause mortality in centenarians are independent of kidney function. Measuring serum TLCκ, TLCλ, and ΣTLC may be a sensitive and highly effective method for predicting all-cause mortality in centenarians. Providing a comprehensive perspective on the production and breakdown processes of immunoglobulins in the body, ΣTLC, in particular, not only correlated linearly with centenarian all-cause mortality, but also had the highest HR value. ΣTLC may also more accurately reflect an individual’s immune status, thereby identifying patients at high risk of death. These findings suggest that monitoring serum TLCκ, TLCλ, and ΣTLC levels may help to identify high-risk individuals in a timely manner, thereby facilitating more detailed further investigation and early intervention.
This study has several strengths. First, it was conducted in a rare cohort of centenarians, who serve as a model of successful aging, with a median follow-up duration of 30 months. Second, this study is likely the first to investigate the relationships between serum TLCκ, TLCλ, and ΣTLC and the risk of death in a large prospective cohort of centenarians. Our findings suggest that serum TLCκ, TLCλ, and ΣTLC can be used independently to predict the risk of death in centenarians. However, further research may be needed to determine whether these biomarkers can be used to extend healthspan. This study also has some limitations. First, the study population was predominantly of Han ethnicity (88.7%), with a small proportion of other ethnic groups (11.3%), and thus our findings may not be generalizable to more diverse ethnic groups or younger populations. Second, this article has not adjusted the healthcare access and socioeconomic status, which may affect the relationship between TLC and all-cause mortality in centenarians. Third, FLC was not detected in this study and will be added in future studies. Finally, the study only examined TLC and all-cause mortality, and further research is needed to determine the relationship between TLC and specific causes of death.
In conclusion, we found that changes in TLC levels are associated with the mortality rate of centenarians. These findings highlight the potential of serum TLCκ, TLCλ, and ΣTLC levels as biomarkers for predicting the mortality rate of centenarians. Future studies will be needed to further clarify the underlying mechanisms and explore the clinical application value of these biomarkers in the health management of elderly individuals.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
Thank all the participants.
Author contributions
Conception and design: Yuting Duan, Weiguang Zhang and Yizhi Chen; Methodology: Zhe Li and Yue Niu; Data acquisition: Bin Wang, Hongyan Hu; Data analysis and interpretation: Yuting Duan, Zhe Feng, Ding Sun, Qiushi Wang, Xinye Jin, Jie Zhang. Drafting of the first draft: Yuting Duan, Yizhi Chen; Writing, review, and/or revision of the manuscript: Song Hu, Hao Li, Yali Zhao, Zehao Zhang, Zeyu Qu, Guangyan Cai, Miao Liu and Yao He; Supervision: Xiangmei Chen. Yuting Duan, Zhe Li and Weiguang Zhang shared the co-first author; Song Hu, Xiangmei Chen and Yizhi Chen shared the co-corresponding author.
Funding
This study was supported by the National Natural Science Foundation of China (No. 82270769, 82173589, 82173590), the Sanya science and technology innovation special project (No. 2022KJCX02), the Specific research fund of the Innovation Platform for Academicians of Hainan Province, Beijing Natural Science Foundation (No.7242033), National Key Research and Development Program of China (No.2022YFC3602900, 2022YFC3602902, 2022YFC3602903, and 2022YFC2503605), Capital’s Funds for Health Improvement and Research (No. CFH 2024-1-5021), Science & Technology Project of Beijing (No. Z221100007422121).
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethical approval
The study was approved by the Ethics Committee of the Chinese People’s Liberation Army (PLA) General Hospital (No. 301HNLL-2016-01), and was conducted in accordance with the Declaration of Helsinki and its subsequent revisions. All the participants provided signed informed content and agreed to attend this survey.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yuting Duan, Zhe Li and Weiguang Zhang contributed equally to this work.
Contributor Information
Song Hu, Email: husong@qdu.edu.cn.
Xiangmei Chen, Email: xmchen301@126.com.
Yizhi Chen, Email: yizchen@126.com.
References
- 1.Cai Y, Song W, LI J, et al. The landscape of aging [J]. Sci China Life Sci. 2022;65(12):2354–454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Franceschi C, Monti D, Sansoni P, et al. The immunology of exceptional individuals: the lesson of centenarians [J]. Immunol Today. 1995;16(1):12–6. [DOI] [PubMed] [Google Scholar]
- 3.Rhyne R L Waynesj, Garry PJ, et al. Cell-mediated immunity as a predictor of morbidity and mortality in subjects over 60 [J]. J Gerontol. 1990;45(2):M45–8. [DOI] [PubMed] [Google Scholar]
- 4.Zhai L, Zhao Y. Detection of the value of consecutive serum total light chain (sTLC) in patients diagnosed with diffuse large B cell lymphoma [J]. Ann Hematol. 2016;95(12):1999–2007. [DOI] [PubMed] [Google Scholar]
- 5.BergóN E, Miravalles E, BergóN E, et al. The predictive power of serum kappa/lambda ratios for discrimination between monoclonal gammopathy of undetermined significance and multiple myeloma [J]. Clin Chem Lab Med. 2005;43(1):32–7. [DOI] [PubMed] [Google Scholar]
- 6.Kaya Z Z, Tuzuner M B, Sahin B, et al. Kappa/Lambda light-chain typing in alzheimer’s disease [J]. Curr Alzheimer Res. 2022;19(1):84–93. [DOI] [PubMed] [Google Scholar]
- 7.He Y, Zhao Y, Yao Y, et al. Cohort profile: the China Hainan centenarian cohort study (CHCCS) [J]. Int J Epidemiol. 2018;47(3):694–h5. [DOI] [PubMed] [Google Scholar]
- 8.Zhang W, Duan Y, LI Z, et al. Association between serum IgM and all-cause mortality risk in Chinese centenarians: a prospective cohort study [J]. Immun Ageing. 2024;21(1):70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Alberti K G, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation [J]. Diabet Med. 1998;15(7):539–53. [DOI] [PubMed] [Google Scholar]
- 10.Levey A S, Stevens L A, Schmid C H, et al. A new equation to estimate glomerular filtration rate [J]. Ann Intern Med. 2009;150(9):604–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schober P, Boer C. Correlation coefficients: appropriate use and interpretation [J]. Anesth Analg. 2018;126(5):1763–8. [DOI] [PubMed] [Google Scholar]
- 12.Gauthier J, Wu Q V, Gooley TA. Cubic splines to model relationships between continuous variables and outcomes: a guide for clinicians [J]. Bone Marrow Transpl. 2020;55(4):675–80. [DOI] [PubMed] [Google Scholar]
- 13.Vanderweele TJ. Sensitivity analysis in observational research: introducing the E-Value [J]. Ann Intern Med. 2017;167(4):268–74. [DOI] [PubMed] [Google Scholar]
- 14.Nowrousian M R, Brandhorst D, Sammet C, et al. Serum free light chain analysis and urine immunofixation electrophoresis in patients with multiple myeloma [J]. Clin Cancer Res. 2005;11(24 Pt 1):8706–14. [DOI] [PubMed] [Google Scholar]
- 15.Maurer MJ, Micallef I N Cerhanjr, et al. Elevated serum free light chains are associated with event-free and overall survival in two independent cohorts of patients with diffuse large B-cell lymphoma [J]. J Clin Oncol. 2011;29(12):1620–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kumar S, Dispenzieri A, Katzmann JA, et al. Serum Immunoglobulin free light-chain measurement in primary amyloidosis: prognostic value and correlations with clinical features [J]. Blood. 2010;116(24):5126–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Giles HV. Karunanithi K. Performance characteristics and limitations of the available assays for the detection and quantitation of monoclonal free light chains and new emerging methodologies [J]. Antibodies (Basel), 2024, 13(1). [DOI] [PMC free article] [PubMed]
- 18.Matsumori A. Novel biomarkers of inflammation for the management of diabetes: Immunoglobulin-Free light chains [J]. Biomedicines, 2022, 10(3). [DOI] [PMC free article] [PubMed]
- 19.Hutchison C A, Harding S, Hewins P, et al. Quantitative assessment of serum and urinary polyclonal free light chains in patients with chronic kidney disease [J]. Clin J Am Soc Nephrol. 2008;3(6):1684–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kraneveld A D, Kool M, Van Houwelingen A H, et al. Elicitation of allergic asthma by Immunoglobulin free light chains [J]. Proc Natl Acad Sci U S A. 2005;102(5):1578–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Aggarwal R, Sequeira W, Kokebie R, et al. Serum free light chains as biomarkers for systemic lupus erythematosus disease activity [J]. Arthritis Care Res (Hoboken). 2011;63(6):891–8. [DOI] [PubMed] [Google Scholar]
- 22.Jin Y, Savage N M Bollagrj, et al. Light chain multiple myeloma: high serum free light chain concentrations portend renal damage and poorer survival [J]. J Appl Lab Med. 2021;6(6):1592–600. [DOI] [PubMed] [Google Scholar]
- 23.Assi L K, Mcintyre N. The association between polyclonal combined serum free light chain concentration and mortality in individuals with early chronic kidney disease [J]. PLoS ONE. 2015;10(7):e0129980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Dispenzieri A, Katzmann J A, Kyle R A, et al. Use of nonclonal serum Immunoglobulin free light chains to predict overall survival in the general population [J]. Mayo Clin Proc. 2012;87(6):517–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Anandram S, Assi L K, Lovatt T, et al. Elevated, combined serum free light chain levels and increased mortality: a 5-year follow-up, UK study [J]. J Clin Pathol. 2012;65(11):1036–42. [DOI] [PubMed] [Google Scholar]
- 26.Farnsworth C W, Spears G Roemmichb. Clinical specificity of two assays for Immunoglobulin kappa and lambda free light chains [J]. Clin Chem Lab Med. 2024;62(5):929–38. [DOI] [PubMed] [Google Scholar]
- 27.Fraser S D S, Fenton A, Harris S, et al. The association of serum free light chains with mortality and progression to End-Stage renal disease in chronic kidney disease: systematic review and individual patient data Meta-analysis [J]. Mayo Clin Proc. 2017;92(11):1671–81. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.