Participant characteristics
A total of 600 Han Chinese patients over the age of 65 who underwent unilateral total knee arthroplasty were included in the PNDABLE study from January 2018 to January 2020. The reasons for dropping out are shown in Fig. 1. And 545 patients (n = 545) remained for analyses. We found the incidence of POD was 9.7% (n = 53 of the 545 patients) via our postoperative assessments. Another 53 non-POD patients were also enrolled in this study (Fig. 1).
In this study, we found patients in the POD group had higher CAM and MDAS scores than the NPOD group. The preoperative MMSE score showed no significant difference between the POD group [28(26–29)] and the NPOD group [28(27-29.5), P = 0.330]. Postoperatively, the VAS score did not differ between patients with delirium 2(1–3) and without delirium [2(1–3), P = 0.080]. The demographic and clinical data of the participants are summarized in Table 1.
Table 1
Comparison of general condition and surgical condition,CFS biomarkers of unilateral total knee arthroplasty patients .
Variable | POD(N = 53) | Non-POD(N = 53) | P-values |
Age (year) ( mean ± SD) | 73.87 ± 6.91 | 70.43 ± 5.65 | 0.006 |
Gender (female/male) | 22/31 | 20/33 | 0.421 |
Body mass index (kg.m− 2 ) ( mean ± SD) | 24.8 ± 3.6 | 25.7 ± 3.4 | 0.187 |
Education level (year) (median and 25–75 percentile) | 9(6-13.5) | 12(9–14) | 0.326 |
ASA physical status (I/II) | 27/26 | 28/25 | 0.846 |
APOE ε4 carriers (%) | 7(13) | 9(17) | 0.587 |
Preoperative serum cholesterol( mean ± SD) | 4.89 ± 0.97 | 4.62 ± 0.97 | 0.398 |
Preoperative serum high density lipoprotein( mean ± SD) | 1.16 ± 0.23 | 1.15 ± 0.23 | 0.798 |
Preoperative serum low density lipoprotein( mean ± SD) | 2.93 ± 0.65 | 2.72 ± 0.58 | 0.200 |
Preoperative CFS Aβ1−42 (pg•ml− 1 ) ( mean ± SD) | 233.98 ± 135.49 | 300.57 ± 99.39 | 0.013 |
Preoperative CFS Aβ1−40 (pg•ml− 1) ( mean ± SD) | 3437.85 ± 2028.87 | 2645.54 ± 924.07 | 0.011 |
Preoperative CFS T-tau (pg•ml− 1) ( mean ± SD) | 314.49 ± 206.41 | 119.33 ± 55.67 | 0.003 |
Preoperative CFS P-tau (pg•ml− 1) ( mean ± SD) | 130.47 ± 51.15 | 69.03 ± 29.01 | 0.001 |
Preoperative CFS Aβ1−42 / T-tau (median and 25–75 percentile) | 0.64(0.22–2.62) | 2.99(1.66–4.13) | 0.001 |
Preoperative CFS Aβ1−42 /P-tau (median and 25–75 percentile) | 1.44(0.63–4.53) | 4.79(3.04–6.88) | 0.001 |
Preoperative CFS Aβ1−40 / T-tau (median and 25–75 percentile) | 11.49(3.57–29.79) | 25.06(17.18–35.27) | 0.002 |
Preoperative CFS Aβ1−40 /P-tau (median and 25–75 percentile) | 18.56(10.56–54.89) | 36.25(24.62–68.35) | 0.003 |
Preoperative CFS PGRN (pg•ml− 1) ( mean ± SD) | 2717.23 ± 873.37 | 3749.06 ± 1004.45 | 0.001 |
Preoperative MMSE scores (median and 25–75 percentile) | 28(26–29) | 28(27-29.5) | 0.330 |
Time from injury to operation (h) ( mean ± SD) | 48.77 ± 9.54 | 52.13 ± 10.45 | 0.087 |
Duration of anesthesia (min) ( mean ± SD) | 133.97 ± 26.5 | 141.25 ± 30.1 | 0.963 |
Duration of surgery (min) ( mean ± SD) | 125.72 ± 25.13 | 129.47 ± 26.32 | 0.455 |
Transfusion amount (ml) ( mean ± SD) | 582.44 ± 148.65 | 603.91 ± 152.77 | 0.465 |
Postoperative the highest CAM score ( mean ± SD) | 31.81 ± 6.18 | 14.4 ± 2.66 | 0.001 |
Postoperative the highest MDAS score( mean ± SD) | 22.75 ± 5.02 | 5.62 ± 2.43 | 0.001 |
Postoperative the highest VAS score (median and 25–75 percentile) | 2(1–3) | 2(1–3) | 0.080 |
The categorical variables were expressed as counts. Normal data are given as mean ± SD, whereas non-normal data are expressed as median and 25–75 percentile. Abbreviations: POD, postoperative delirium; MMSE, mini-mental state examination; ASA, American Society of Anesthesiologists; MDAS, memorial delirium assessment scale;VAS, Visual Analogue Scale/Score; SD, standard deviation; CSF, cerebrospinal fluid; Aβ1−42, amyloid-β1–42; Aβ1–40, amyloid-β1–40; T-tau, total tau; P-tau, phosphorylated tau; PGRN, Progranulin |
CSF PGRN Concentration
In this study, there were significant differences in preoperative CSF PGRN concentration between POD and NPOD groups (P < 0.001, Table 1). Besides, univariate logistic analysis (adjusted for age, gender, years of education, and APOEε4 carrier status) showed that PGRN was an independent risk factor for POD in elderly patients undergoing unilateral total knee arthroplasty (OR = 1.001, 95% CI 1.001–1.002, P < 0.001, Table 2).
Table 2
Adjusted by gender, age, educational level and APOE ε4
| Unadjusted | Adjusted |
Adjusted odds ratio (95% CI) | P-value | Adjusted odds ratio (95% CI) | P-value |
Preoperative CFS PGRN | 0.999(0.998–0.999) | 0.001 | 0.998(0.997–0.999) | 0.001 |
Preoperative CFS Aβ1−40 | 1.000(1.000-1.001) | 0.017 | 1.001(1.000-1.001) | 0.001 |
Preoperative CFS Aβ1−42 | 0.995( 0.992–0.999) | 0.006 | 0.995( 0.991–0.999) | 0.009 |
Preoperative CFS T-tau | 1.012(1.007–1.018) | 0.001 | 1.012(1.009–1.022) | 0.001 |
Preoperative CFS P-tau | 1.033(1.020–1.047) | 0.001 | 1.043(1.027–1.060) | 0.001 |
Preoperative CFS Aβ1−42 / T-tau | 0.660(0.523–0.832) | 0.001 | 0.633(0.490–0.817) | 0.001 |
Preoperative CFSAβ1−42 /P-tau | 0.776(0.672–0.895) | 0.001 | 0.765(0.654–0.896) | 0.001 |
Preoperative CFS Aβ1−40 / T-tau | 0.998(0.984–1.012) | 0.750 | 1.001(0.985–1.016) | 0.944 |
Preoperative CFS Aβ1−40 /P-tau | 0.996(0.989–1.004) | 0.359 | 0.998(0.990–1.007) | 0.695 |
Since age is the main risk factor for POD, we explored whether CSF PGRN levels were related to aging. We found PGRN levels increased with age, as demonstrated by a significantly positive correlation (r = 0.796, P < 0.001). The results indicated that CSF PGRN did differ significantly between different age subgroups (65–70 years: 2189 ± 119 pg/ml, n = 26, P = 0.001; 65–70 years: 5206 ± 408 pg/ml, n = 18, P = 0.001; >80 years: 7563 ± 502 pg/ml, n = 9; P = 0.037) (Fig. 2).
We found PGRN levels increased with CAM and MDAS scores, as demonstrated by significantly positive correlations (r = 0.781, P < 0.001; r = 0.524, P < 0.001) (Fig. 3).
Differences in CSF PGRN level between different subgroups stratified by biomarkers
The associations between CSF PGRN and CSF core biomarkers for POD were tested in linear regression models adjusted for age, gender, years of education and APOE ε4 carrier status. In the whole sample of subjects (n = 659), increased CSF PGRN was associated with lower levels of Aβ1−42 (β = -0.644, P < 0.001), Aβ1−40 (β = 0.275, P = 0.017), Aβ42/p-tau (β = -0.035, P < 0.001) and Aβ42/T- tau (β = -0.073, P < 0.001), as well as higher levels of T-tau (β = 0.557, P < 0.001) and P-tau (β = 0.502, P < 0.001) (Fig. 4).
There were positive associations of CSF PGRN with T-tau (β = -0.501, P < 0.001) and P-tau (β = -0.470, P < 0.001) and negative associations of CSF PGRN with Aβ1−42 (β = -0.576, P < 0.001), Aβ1−40 (β = -0.488, P < 0.001), Aβ42/p-tau (β = -0.422, P < 0.001) and Aβ42/T-tau (β = -0.395, P < 0.001) in POD patients (n = 53) (Fig. 5).
In the NPOD group (n = 53), we found a positive association between CSF PGRN and T-tau (β = 0.495, P < 0.001) and negative associations of CSF PGRN with Aβ1−42 (β = -0.364, P < 0.001), Aβ42/p-tau (β = -0.019, P < 0.001) and Aβ42/T-tau (β = -0.050, P < 0.001), whereas no significant association was found of CSF sTREM2 with p-tau (β = 0.181,P = 0.069) or Aβ1−40 (β = 0.001, P = 0.998) (Fig. 6).
We then calculated the ratios between CSF amyloid and tau biomarkers, and found no associations of CSF PGRN with CSF Aβ1−40/T-tau or Aβ1−40/p-tau (Table 2). Outliers were excluded in our analyses, but we obtained similar results when those were included. These findings indicate that higher CSF PGRN correlates with lower levels of Aβ and higher levels of tau.
Receiver operating characteristic (ROC) curve analysis of PGRN in CSF
The ROC curve analysis of PGRN showed that PGRN concentration had high diagnostic value for POD, with all the AUC greater than 0.5 and close to 1.0. (Table 3, Fig. 7).
Table 3
The ROC curve analysis of PGRN showed that the concentrations of CSF,s PGRN had high diagnostic value for POD
CSF,s index | AUC | 95%CI(L) | 95%CI(U) | Youden,s index | Sensitivity | Specificity |
PGRN ((pg•ml− 1) | 0.795 | 0.706 | 0.867 | 0.528 | 58.5 | 94.3 |