Volume and Body Composition in Hemodialysis Patients: A Bioimpedance Study Assessing Differences by Sex

Individuals with kidney disease receiving maintenance hemodialysis exhibit improved survival with higher baseline body mass index and weight gain over time (1Kalantar-Zadeh K. Kopple J.D. Kilpatrick R.D. McAllister C.J. Shinaberger C.S. Gjertson D.W. et al.Association of Morbid Obesity and Weight Change Over Time With Cardiovascular Survival in Hemodialysis Population.Am J Kidney Dis. 2005 Sep; 46: 489-500Abstract Full Text Full Text PDF PubMed Scopus (253) Google Scholar). This reversal of risk factors compared to the general population (often misnamed "reverse epidemiology" (2Levin N.W. Handelman G.J. Coresh J. Port F.K. Kaysen G.A. Reverse Epidemiology: A Confusing, Confounding, and Inaccurate Term.Semin Dial. 2007 Nov; 20: 586-592Crossref PubMed Scopus (64) Google Scholar)) has been substantiated by metrics of body composition obtained through bioimpedance spectroscopy. Specifically, bioimpedance-spectroscopy-data have indicated survival was optimal with pre-dialysis lean tissue index (LTI) ranging from 15-20kg/m2 and fat tissue index (FTI) from 4-15kg/m2, respectively (3Marcelli D. Usvyat L.A. Kotanko P. Bayh I. Canaud B. Etter M. et al.Body composition and survival in dialysis patients: results from an international cohort study.Clin J Am Soc Nephrol CJASN. 2015 Jul 7; 10: 1192-1200Crossref PubMed Scopus (0) Google Scholar). Bioimpedance-derived estimates of pre-dialysis fluid status are also associated with increased mortality when exceeding 2.5L absolute volume overload (VO) and 15% relative VO as a percentage of extracellular water (4Wizemann V. Wabel P. Chamney P. Zaluska W. Moissl U. Rode C. et al.The mortality risk of overhydration in haemodialysis patients.Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc - Eur Ren Assoc. 2009 May; 24: 1574-1579Crossref PubMed Scopus (0) Google Scholar). Although reference individuals for LTI and FTI were age- and sex-matched (5Wieskotten S. Heinke S. Wabel P. Moissl U. Becker J. Pirlich M. et al.Bioimpedance-based identification of malnutrition using fuzzy logic.Physiol Meas. 2008 May 1; 29: 639-654Crossref PubMed Scopus (62) Google Scholar), those for VO were not (4Wizemann V. Wabel P. Chamney P. Zaluska W. Moissl U. Rode C. et al.The mortality risk of overhydration in haemodialysis patients.Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc - Eur Ren Assoc. 2009 May; 24: 1574-1579Crossref PubMed Scopus (0) Google Scholar). In light of current efforts to better understand sex discrepancies in kidney disease patients (6Chesnaye N.C. Carrero J.J. Hecking M. Jager K.J. Differences in the epidemiology, management and outcomes of kidney disease in men and women.Nat Rev Nephrol. 2024 Jan; 20: 7-20Crossref Scopus (7) Google Scholar), we aimed to explore whether there are differences in pre-dialysis body composition and fluid status between male and female patients on hemodialysis. We retrospectively analyzed pre-dialysis bioimpedance measurements conducted in patients on maintenance hemodialysis between November 2022 and January 2023 at the "Vienna Dialysis-Center" (a tertiary care facility in Vienna, Austria). Patients were measured with the Body Composition Monitor (Fresenius Medical Care) and the Cella (Cella Medical) bioimpedance spectroscopy devices in standard wrist-to-ankle setups using pre-gelled electrodes. Detailed methodology and further results are provided in the Item S1-S2. In November 2022, the Vienna Dialysis-Center cared for 304 patients (285 patients on regular twice- or thrice-weekly hemodialysis, 11 undergoing in-hospital treatment, 8 on vacation). Pre-dialysis bioimpedance data from 159 patients were available for initial evaluation. After excluding erroneous measurements (Fig S1), 137 patients remained for sex comparisons, 85 (62.0%) of whom were male and 52 (38.0%) were female. Study population data and bioimpedance data from the Body Composition Monitor are shown overall and stratified by sex in Table 1. Pre-dialysis VO (higher in males), LTI (higher in males) and FTI (higher in females) differed significantly between sexes while relative VO did not. The proportion of patients within LTI measurements of 15-20kg/m2 (the range with optimal survival) differed significantly between males and females (27.1% vs. 7.7%, p<0.01), but there was no difference for the above-mentioned thresholds of absolute and relative VO or FTI (Fig 1 and Fig S2). Body Composition Monitor and Cella devices differed significantly in all parameters (Table S1 and Fig S3-S7).Table 1General data, dialysis therapy data and bioimpedance data of the study population overall and stratified by sexCharacteristicMissing (n, %)Overall, n = 137Male, n = 85Female, n = 52PPopulation characteristicsAge, years0 (0%)63.0 (52.0, 74.0)60.0 (48.0, 72.0)65.5 (57.8, 75.3)0.08 †Height, cm0 (0%)170.0 (162.0, 176.0)174.0 (170.0, 180.0)160.0 (157.0, 164.3)<0.01 †Body mass index, kg/m20 (0%)26.6 (23.2, 31.1)26.5 (23.3, 30.6)27.1 (22.4, 33.3)0.6 †Body mass pre-HD, kg2 (1.5%)78.3 (67.1, 89.6)80.8 (69.6, 94.4)72.7 (63.3, 86.6)<0.01 †Intradialytic weight loss, kg5 (3.6%)2.4 (1.3, 3.0)2.5 (1.8, 3.3)2.0 (1.1, 2.5)<0.01 †Difference to dry weight post-HD, kg10 (7.3%)0.1 (-0.1, 0.8)0.1 (-0.1, 1.0)0.1 (0.0, 0.7)0.9 †Interdialytic weight gain from last session, kg4 (2.9%)2.3 (1.1, 3.2)2.5 (1.6, 3.3)1.8 (0.7, 2.5)<0.01 †Interdialytic weight gain to next session, kg6 (4.4%)2.3 (1.0, 3.1)2.5 (1.4, 3.3)1.8 (0.5, 2.8)0.01 †Dialysis vintage, months3 (2.2%)27.8 (12.6, 56.8)28.9 (12.1, 52.7)27.2 (14.8, 61.9)0.8 †Ultrafiltration volume, L4 (2.9%)2.5 (1.7, 3.1)2.8 (1.9, 3.5)2.3 (1.5, 2.7)<0.01 †Dialysis therapy duration, minutes43 (31%)228.5 (207.3, 234.0)229.0 (218.0, 234.0)212.0 (181.3, 234.0)0.03 †Systolic BP pre-HD, mmHg3 (2.2%)144.5 (131.3, 158.0)143.0 (131.0, 158.0)146.0 (135.3, 157.0)0.8 †Diastolic BP pre-HD, mmHg3 (2.2%)73.5 (65.0, 82.0)74.5 (67.0, 83.3)71.5 (62.3, 81.0)0.12 †Systolic BP post-HD, mmHg4 (2.9%)140.0 (123.0, 156.0)138.0 (120.0, 157.5)144.0 (126.5, 153.8)0.7 †Diastolic BP post-HD, mmHg4 (2.9%)72.0 (63.0, 82.0)74.0 (64.0, 86.5)70.0 (62.0, 77.0)0.02 †Intradialytic hypotension, (n, %)7 (5.1%)13.0 (10.0%)7.0 (8.8%)6.0 (12.0%)0.8 ‡Venous hemoglobin pre-HD, g/dL3 (2.2%)11.4 (10.7, 12.2)11.7 (10.8, 12.3)11.2 (10.6, 12.0)0.09 †Previous albumin, g/dL2 (1.5%)3.4 (3.2, 3.7)3.4 (3.2, 3.7)3.4 (3.2, 3.7)0.7 †Previous total protein, g/dL2 (1.5%)6.9 (6.5, 7.2)7.0 (6.6, 7.2)6.9 (6.4, 7.2)0.2 †Bioimpedance spectroscopy characteristicsExtracellular resistance, Ohm0 (0%)530.2 (457.9, 590.6)518.5 (446.4, 564.1)554.4 (494.6, 620.6)0.01 †Intracellular resistance, Ohm0 (0%)1,596.7 (1,337.0, 1,959.3)1,501.7 (1,247.7, 1,838.5)1,844.2 (1,406.1, 2,174.0)<0.01 †Resistance at infinite frequency, Ohm0 (0%)400.8 (342.3, 445.6)386.1 (327.4, 425.5)427.4 (371.0, 476.4)<0.01 †Capacitance, nF0 (0%)1.2 (0.9, 1.7)1.4 (1.0, 1.8)1.0 (0.8, 1.3)<0.01 †Time delay, ns0 (0%)0.2 (-1.3, 3.1)0.3 (-1.4, 3.1)0.2 (-1.3, 3.3)0.7 †α, radian0 (0%)0.6 (0.6, 0.7)0.7 (0.6, 0.7)0.6 (0.6, 0.7)0.04 †Extracellular water, L0 (0%)18.5 (16.3, 21.3)19.8 (17.9, 22.7)16.3 (14.1, 18.3)<0.01 †Intracellular water, L0 (0%)18.4 (15.6, 21.5)20.7 (17.7, 23.3)15.6 (13.6, 18.0)<0.01 †Total body water, L0 (0%)36.7 (31.8, 42.6)40.7 (36.2, 45.6)31.4 (28.1, 35.8)<0.01 †Volume overload pre-HD, L0 (0%)2.4 (1.4, 3.5)2.7 (1.5, 4.0)1.9 (1.0, 3.0)0.03 †Volume overload pre-HD <2.5 L0 (0%)69.0 (50.4%)37.0 (43.5%)32.0 (61.5%)0.06 ‡Relative volume overload pre-HD, %0 (0%)13.4 (7.9, 18.5)14.3 (8.2, 19.5)12.0 (6.0, 17.3)0.4 †Relative volume overload pre-HD <15%0 (0%)80.0 (58.4%)47.0 (55.3%)33.0 (63.5%)0.4 ‡Lean tissue index, kg/m20 (0%)12.4 (10.5, 14.6)13.4 (11.2, 15.6)11.0 (9.5, 13.0)<0.01 †Lean tissue index Δ reference, kg/m20 (0%)0.0 (-1.9, 1.7)-0.4 (-2.1, 1.7)0.4 (-0.8, 2.1)0.02 †Lean tissue index between 15-20 kg/m20 (0%)27.0 (19.7%)23.0 (27.1%)4.0 (7.7%)<0.01 ⋇Fat tissue index, kg/m21 (0.7%)12.2 (9.6, 17.7)11.1 (9.1, 16.4)14.2 (9.9, 22.0)0.04 †Fat tissue index Δ reference, kg/m21 (0.7%)6.5 (3.6, 11.3)6.4 (3.7, 9.3)7.1 (3.0, 14.6)0.5 †Fat tissue index between 4-15 kg/m21 (0.7%)83.0 (61.0%)57.0 (67.9%)26.0 (50.0%)0.06 ‡Characteristics are reported as number (%) or median (interquartile range). Albumin and total protein measurements were extracted form the most recent routine laboratory before the bioimpedance measurement. Abbreviations: BP, blood pressure; HD, hemodialysis.†Two-sample Wilcoxon test‡Chi-square test⋇Fisher's exact test Open table in a new tab Characteristics are reported as number (%) or median (interquartile range). Albumin and total protein measurements were extracted form the most recent routine laboratory before the bioimpedance measurement. Abbreviations: BP, blood pressure; HD, hemodialysis. †Two-sample Wilcoxon test ‡Chi-square test ⋇Fisher's exact test In summary, we found significant differences between males and females in pre-dialysis VO, LTI and FTI not only in absolute terms, but also in the proportion of patients considered to be "in-range" of LTI. Although higher LTI was previously associated with lower mortality in patients on maintenance hemodialysis (3Marcelli D. Usvyat L.A. Kotanko P. Bayh I. Canaud B. Etter M. et al.Body composition and survival in dialysis patients: results from an international cohort study.Clin J Am Soc Nephrol CJASN. 2015 Jul 7; 10: 1192-1200Crossref PubMed Scopus (0) Google Scholar), bioimpedance-derived sarcopenia was not (Item S3a). In another study, subjectively-assessed muscle atrophy was associated with mortality more often in females, who also exhibited poor nutritional status (Item S3b). In our study, differences in pre-dialysis VO were present only in absolute terms. Other studies, while not prominently stressing these findings, also indicate larger absolute (Item S3c-e) and even relative (Item S3f) pre-dialysis VO in males. While some authors suggested that the threshold for relative VO should be sex-specific (7Zoccali C. Moissl U. Chazot C. Mallamaci F. Tripepi G. Arkossy O. et al.Chronic Fluid Overload and Mortality in ESRD.J Am Soc Nephrol JASN. 2017 Aug; 28: 2491-2497Crossref PubMed Scopus (0) Google Scholar), evidence supporting this claim is so far lacking. Reasons for pre-dialysis differences may in part originate from physiological differences (less extracellular water, therefore less VO) and in part from varying patient preferences. Differences in body composition between sexes are expected, but whether differences in VO are true remains unclear. In bioimpedance spectroscopy, body composition is modeled using the input variables extra- and intracellular resistance, body mass and body height (8Moissl U.M. Wabel P. Chamney P.W. Bosaeus I. Levin N.W. Bosy-Westphal A. et al.Body fluid volume determination via body composition spectroscopy in health and disease.Physiol Meas. 2006 Sep; 27: 921-933Crossref PubMed Scopus (500) Google Scholar,9Chamney P.W. Wabel P. Moissl U.M. Müller M.J. Bosy-Westphal A. Korth O. et al.A whole-body model to distinguish excess fluid from the hydration of major body tissues.Am J Clin Nutr. 2007 Jan; 85: 80-89Abstract Full Text Full Text PDF PubMed Scopus (390) Google Scholar). In the equations used in the Body Composition Monitor, sex-dependent resistivities, body shape factor and body density were combined into one empirically-derived expression with linear dependence on body mass index (8Moissl U.M. Wabel P. Chamney P.W. Bosaeus I. Levin N.W. Bosy-Westphal A. et al.Body fluid volume determination via body composition spectroscopy in health and disease.Physiol Meas. 2006 Sep; 27: 921-933Crossref PubMed Scopus (500) Google Scholar). Reference populations to derive this body composition model did not exhibit under- or overrepresentation of sexes (8Moissl U.M. Wabel P. Chamney P.W. Bosaeus I. Levin N.W. Bosy-Westphal A. et al.Body fluid volume determination via body composition spectroscopy in health and disease.Physiol Meas. 2006 Sep; 27: 921-933Crossref PubMed Scopus (500) Google Scholar,9Chamney P.W. Wabel P. Moissl U.M. Müller M.J. Bosy-Westphal A. Korth O. et al.A whole-body model to distinguish excess fluid from the hydration of major body tissues.Am J Clin Nutr. 2007 Jan; 85: 80-89Abstract Full Text Full Text PDF PubMed Scopus (390) Google Scholar) but were based on narrower ranges of body mass index than typical hemodialysis populations (25.2 ± 3.7kg/m2 (9Chamney P.W. Wabel P. Moissl U.M. Müller M.J. Bosy-Westphal A. Korth O. et al.A whole-body model to distinguish excess fluid from the hydration of major body tissues.Am J Clin Nutr. 2007 Jan; 85: 80-89Abstract Full Text Full Text PDF PubMed Scopus (390) Google Scholar) vs. 26 ± 5.3kg/m2 (3Marcelli D. Usvyat L.A. Kotanko P. Bayh I. Canaud B. Etter M. et al.Body composition and survival in dialysis patients: results from an international cohort study.Clin J Am Soc Nephrol CJASN. 2015 Jul 7; 10: 1192-1200Crossref PubMed Scopus (0) Google Scholar)), which could lead to erroneous estimates in patients outside of the initial validation range. Body shape is largely different between sexes, independent of body mass index. The present study is limited by its relatively small sample size and retrospective, cross-sectional design. In conclusion, we identified notable sex differences in pre-dialysis fluid status and body composition among patients on maintenance hemodialysis at a single center. For VO, these differences have also been observed but have not been commented on or discussed in previous datasets. Considering that the survival of females undergoing hemodialysis is equal to (although in some strata slightly better than) males (10ERA Registry: ERA Registry Annual Report 2021. Amsterdam UMC, location AMC, Department of Medical Informatics, Amsterdam, the Netherlands, 2023.Google Scholar), thresholds of body composition parameters associated with increased mortality may not be equal between sexes if the values themselves are sex-dependent. We therefore recommend reanalyzing existing large-scale datasets for sex differences in mortality and bioimpedance-derived measures of body composition. Conceptualization: SM, PW, MH; Methodology: SM; Software: SM; Formal Analysis: SM; Investigation: SM, JN, CM, SW; Resources: MH; Data curation: SM; Visualization: SM; Supervision: MH; Project administration: SM, MH; Funding acquisition: MH. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. This work was supported by the Vienna Science and Technology (WWTF) Precision Medicine Grant LS20-079. CC and PW were previously employees of Fresenius Medical Care, which produces the Body Composition Monitor. DS is a coinventor of patents in the field of blood volume and bioimpedance applications in hemodialysis and is a member of the American Renal Associates research board. MH served as a speaker and/or consultant for Astellas Pharma, AstraZeneca, Eli Lilly, Fresenius Medical Care, Janssen-Cilag, Siemens Healthcare, and Vifor and has previously received academic study support from Astellas Pharma, Boehringer Ingelheim, Eli Lilly, Nikkiso, and Siemens Healthcare (not related to the present work). The remaining authors have nothing to declare. Data will be made available upon reasonable request to the corresponding author. Received March 2, 2024. Evaluated by 1 external peer reviewer, with direct editorial input from the Statistical Editor and the Editor-in-Chief. Accepted in revised form April 30, 2024. The authors wish to thank all medical staff at the Vienna Dialysis-Center and Ole Schecker and Moritz Reuth for additional data acquisition. Download .pdf (1.49 MB) Help with pdf files