A pre-operative high-protein diet can improve the serum albumin levels of patients undergoing total knee arthroplasty

To the Editor: Malnutrition is associated with an increased risk of complications after total knee arthroplasty (TKA), including persistent wound drainage, delay of wound healing, superficial surgical site and even deep periprosthetic joint infections, and intensive care unit admission.^[1] Albumin (ALB) is the most abundant protein in human plasma and is considered a reliable and sensitive biomarker of nutritional status and malnutrition. Like malnutrition, low serum ALB levels are not only associated with various post-operative complications following TKA including pulmonary infection, renal impairment, superficial and deep infections, unplanned intubation, and even mortality; but also associated with increased treatment costs, longer length of stay and higher risk of readmission.^[2]

Human albumin solution (HAS) is frequently used to correct hypoproteinemia. However, a study including >1 million patients from 510 hospitals in the United States showed that HAS use was associated with an increased risk of thromboembolic, acute renal failure, and intensive care unit admission.^[3] Furthermore, HAS is expensive, placing a substantial financial burden on health care systems.

Pre-operatively correcting hypoalbuminemia and increasing ALB levels may be an effective way to reduce postoperative ALB use due to the elective nature of TKA. Therefore, this study was performed to illustrate the effect of pre-operative ALB levels on human ALB use and to determine the effect of a high-protein diet during the preoperative waiting period on serum ALB levels in patients scheduled to undergo TKA.

This was a single-center retrospective study, which was approved by the Ethics Committee of West China Hospital of Sichuan University (No. 2020–804), and registered in the Chinese Clinical Trial Registry (ChiCTR2000034978). The participants of the study consisted of two cohorts. First, patients (n = 660) who underwent primary TKA due to primary osteoarthritis or rheumatoid arthritis in our hospital from January to December 2019 were enrolled to investigate the effect of pre-operative ALB levels on HAS use. Second, patients (n = 88) who underwent primary TKA from January to July 2020 were included to evaluate the effect of a high-protein diet on pre-operative serum ALB levels; those patients were educated to eat a high-protein diet during the pre-operative waiting period. The criterion for post-operative use of HAS was serum ALB concentration under 35 g/L. Patients were excluded if they had been diagnosed with hemophilia, tuberculous arthritis, liver and/ or kidney disease, or if relevant data about them were missing.

When a patient decided to undergo TKA for severe knee osteoarthritis or rheumatoid arthritis at the outpatient clinic, admission time was booked. The patient's diet and nutritional status were assessed according to the levels of hemoglobin (Hb) and serum ALB. If Hb <140 g/L or ALB <45 g/L, the patients with body mass index (BMI) < 25 kg/m² were informed that they needed to increase the proportion of protein in their diet during the pre-operative waiting period to prepare for the surgery. The recommended daily diet was as follows: at least three eggs (fried eggs were not recommended) and 50 g lean meat, which contains approximately 35 g of protein.

For the 2019 cohort, data were collected on demographic characteristics including sex, age, BMI; clinical characteristics including diagnosis, and comorbidities such as hypertension and diabetes; results of laboratory tests on the day of admission such as Hb, serum ALB, and HAS use. For the 2020 cohort, in addition to the previously mentioned demographic and clinical characteristics, data were also collected during the waiting period before admission and the results of laboratory tests on the day of outpatient visit, including ALB, globulin Hb, hematocrit (HCT), biomarkers of liver and kidney function, and lipid metabolism.

Continuous variables were presented as the mean ± standard deviation, while categorical variables were shown as the frequency (proportion). The relationship between HAS use (treated as a binary variable) and pre-operative serum ALB levels (treated as a continuous variable) was expressed as the relative risk (RR) with a 95% confidence interval (CI). A crude model of HAS use regarding pre-operative serum ALB levels was built using univariate analyses. Logistic regression analyses were used to adjust for other variables. The correlations among the covariates included in the logistic regression were quantified by Spearman's rank correlation analysis. Furthermore, restrictive cubic spline regression analyses were used to detect the dose-response relationships between serum ALB levels tested on the day of admission and post-operative ALB use. The dose response was analyzed with R version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria). The paired-samples t test was used to compare the difference of biomarkers between admission and outpatient visits. P < 0.05 was considered as statistical significance. All data analyses were performed using SPSS version 21 software (IBM, Armonk, NY, USA).

A total of 748 patients were included, of whom 660 patients underwent TKA from January to December 2019 while 88 patients underwent the procedure from January to July 2020, and among them, 4 (4.5%) patients were diagnosed with rheumatoid arthritis of the knee. Among patients from the 2019 cohort, 97 (14.7%) were treated with HAS. For patients recruited in 2020, the mean pre-operative waiting period was 39.05 ± 22.78 days. The baseline demographic and clinical characteristics of the 2019 cohort are shown in Supplementary Table 1, https://links.lww.com/CM9/B97, while those from the 2020 cohort are presented in Supplementary Table 2, https://links.lww.com/CM9/B98.

Spearman's rank correlation analyses showed that the absolute value of the rank correlation coefficient among covariates included in logistic regression analyses ranged from 0 to 0.375, which meant that correlations among the variables were low and that variables met the requirements for logistic regression analyses [Supplementary Figure 1, https://links.lww.com/CM9/B94].

Several statistical models including different covariates were built to evaluate the impact of pre-operative ALB levels at admission on post-operative HAS use in primary TKA [Supplementary Table 3, https://links.lww.com/CM9/B99]. The univariate analysis (crude model) showed that higher pre-operative ALB levels were associated with a lower risk of post-operative HAS use (RR = 0.720, 95%

CI 0.665–0.778, P < 0.001). After adjusting for demographic variables (age, sex, and BMI) and clinical characteristics in the logistic regression analysis step-by-step (models 1, 2, and 3), high pre-operative ALB levels remained significantly associated with a lower risk of HAS use [Supplementary Table 3, https://links.lww.com/CM9/B99]. In addition, a negative dose-response relationship was observed between pre-operative ALB levels and the risk of post-operative HAS use (P_overall < 0.001; P_non-linear = 0.186; Supplementary Figure 2, https://links.lww.com/CM9/B95).

As shown in Table 1, the mean ALB level on the day of admission was significantly higher than the level tested on the day of outpatient visit (46.04 ± 2.49 g/L vs. 42.79 ± 4.09 g/L; P < 0.001), while there were no significant differences between admission and outpatient visits in the levels of globulin, Hb, HCT, and fasting blood glucose, as well as in the biomarkers of liver and kidney function and lipid metabolism (all P > 0.05).

Table 1 - The effect of high-protein diet on different biomarkers in patients undergoing primary TKA from the 2020 cohort.

Biomarkers	Outpatient visit (n = 88)	Pre-op (n = 88)	P values
Nutritional status
ALB (g/L)	42.79 ± 4.09	46.04 ± 2.49	<0.001
Globulin (g/L)	28.29 ± 4.59	27.64 ± 3.81	0.158
Hb (g/L)	134.63 ± 12.59	136.34 ± 11.48	0.069
HCT (L/L)	0.41 ± 0.38	0.42 ± 0.32	0.585
Liver function
Total bilirubin (μmol/L)	10.29 ± 3.98	10.57 ± 3.98	0.569
Direct bilirubin (μmol/L)	3.01 ± 1.49	2.91 ± 1.11	0.496
Indirect bilirubin (μmol/L)	7.24 ± 2.97	7.65 ± 3.14	0.273
Total bile acid (μmol/L)	4.17 ± 2.11	3.67 ± 2.77	0.157
ALT (IU/L)	27.55 ± 20.31	24.08 ± 13.76	0.107
AST (IU/L)	23.21 ± 5.95	22.61 ± 7.65	0.438
ALP (IU/L)	97.17 ± 29.17	93.26 ± 25.48	0.090
GGT (IU/L)	35.45 ± 30.92	33.68 ± 26.85	0.473
Kidney function
Serum urea (μmol/L)	5.69 ± 1.72	5.94 ± 1.78	0.082
Serum creatinine (μmol/L)	66.66 ± 17.58	66.53 ± 16.37	0.876
eGFR (mL ·min–1 1.73 m–2)	88.22 ± 18.82	88.69 ± 17.04	0.493
Serum uric acid (μmol/L)	316.58 ± 103.51	310.40 ± 109.96	0.549
Lipid metabolism
Triglyceride (mmol/L)	1.57 ± 0.60	1.47 ± 0.84	0.261
Cholesterin (mmol/L)	4.88 ± 0.90	5.01 ± 0.96	0.150
HDL-C (mmol/L)	1.35 ± 0.35	1.39 ± 0.34	0.103
LDL-C (mmol/L)	2.82 ± 0.69	2.92 ± 0.69	0.118
Other
Creatine kinase (IU/L)	101.36 ± 63.63	91.56 ± 41.43	0.079
Lactic dehydrogenase (IU/L)	196.59 ± 26.02	192.02 ± 38.81	0.438
HBDH (IU/L)	155.17 ± 22.89	150.65 ± 32.98	0.329
Fasting blood glucose (mmol/L)	5.63 ± 1.26	5.46 ± 0.86	0.208

Data are presented as mean ± standard deviation.ALB: Albumin; ALP: Alkaline phosphatase; ALT: Alanine transferase; AST: Aspartate transferase; eGFR: Estimated glomerular filtration rate; GGT: Glutamyl transpeptidase; Hb: Hemoglobin; HCT: Hematocrit; HDL-C: High-density lipoprotein cholesterol; HBDH: Hydroxybutyrate dehydrogenase; LDL-C: Low-density lipoprotein cholesterol; Pre-op: Preoperative; TKA: Total knee arthroplasty.

Serum ALB levels are associated with long-term mortality in the general population, especially for the elderly, and with post-operative course and mortality for patients with geriatric hip fracture.^[4] In addition, although serum ALB < 35 g/L is usually considered as malnutrition or hypoproteinemia, our study revealed a negative dose-dependent relationship between pre-operative serum ALB levels and post-operative HAS use, consistent with a study showing that lower pre-operative ALB levels are associated with higher treatment costs.^[2]

In our patients, approximately 35 g of dietary protein (approximately 0.5 g/kg body weight) was added to the daily diet, which resulted in an increase of approximately 3.2 g in serum ALB from the day of outpatient visit to admission (an average of 39.05 ± 22.78 days). Several studies have previously evaluated the effects of a high-protein diet. High-protein diet (1.07–1.60 g/kg body weight/day) lasting for 6 to 12 months was an effective and safe way to reduce weight: high protein produced satiety and increased energy expenditure.^[5]

Our study showed that a high-protein diet had no significant impact on biomarkers of liver, kidney, or lipid metabolism. A high-protein diet lasting for approximately 34 months had no significant effect on the biomarkers of cardiometabolic health and vascular function in overweight participants.^[6]

Our study presents several limitations. First, we excluded patients with renal and hepatic diseases and therefore, the safety of a high-protein diet for those patients must be assessed further because of its potential effects on renal function. Second, although we educated patients in detail, some patients may have not strictly followed our suggestions during the pre-operative waiting period.

In conclusion, lower pre-operative serum ALB levels were associated with an increased risk of post-operative HAS use with a negative dose-response relationship in TKA. A high-protein diet during the waiting period before surgery admission may be a useful way to improve ALB levels in these patients.

Supplementary file, https://links.lww.com/CM9/B96

Acknowledgements

The authors thank A. Chapin Rodríguez, PhD, from Creadu-cate Enterprises Ltd. for editing the English text of a draft of this manuscript.

Funding

This study was supported by a grant from the 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (No. ZYJC18039).

Conflicts of interest

None.