3.1 baseline demographics of the study participants
A total of 283 people participated in the analysis, and the baseline demographic characteristics of participants are shown in Table 1.
Table 1
Baseline demographic characteristics of patients in the progression and progression-free group
|
|
Overall
|
progression
|
progression-free
|
p value
|
Group (%)
|
|
283
|
134 (47.35)
|
149 (52.65)
|
< 0.0001
|
Sex (%)
|
Female
|
79 (27.92)
|
35 (26.12)
|
44 (29.53)
|
0.6129
|
Male
|
204 (72.08)
|
99 (73.88)
|
105 (70.47)
|
|
Age (%)
|
< 40y
|
96 (33.92)
|
18 (13.43)
|
78 (52.35)
|
< 0.0001
|
40 ~ 59y
|
91 (32.16)
|
49 (36.57)
|
42 (28.19)
|
|
≥ 60y
|
96 (33.92)
|
67 (50.00)
|
29 (19.46)
|
|
BMI (%)
|
< 18.50
|
79 (27.92)
|
14 (10.45)
|
65 (43.62)
|
< 0.0001
|
18.50 ~ 24.99
|
142 (50.18)
|
81 (60.45)
|
61 (40.94)
|
|
≥ 25.00
|
62 (21.91)
|
39 (29.10)
|
23 (15.44)
|
|
Cough (%)
|
|
|
|
|
0.0918
|
no
|
152 (53.71)
|
64 (47.76)
|
88 (59.06)
|
|
yes
|
130 (45.94)
|
70 (52.24)
|
60 (40.27)
|
|
Dyspnea (%)
|
|
|
|
|
0.3414
|
no
|
186 (65.72)
|
91 (67.91)
|
95 (63.76)
|
|
yes
|
95 (33.57)
|
43 (32.09)
|
52 (34.90)
|
|
Ethnic (%)
|
Han
|
272 (96.11)
|
130 (97.01)
|
142 (95.30)
|
0.7397
|
Qiang
|
1 (0.35)
|
0 (0.00)
|
1 (0.67)
|
|
Tibetan
|
4 (1.41)
|
2 (1.49)
|
2 (1.34)
|
|
Tu
|
1 (0.35)
|
0 (0.00)
|
1 (0.67)
|
|
Uyghur
|
1 (0.35)
|
0 (0.00)
|
1 (0.67)
|
|
Yi
|
4 (1.41)
|
2 (1.49)
|
2 (1.34)
|
|
Smoking history (%)
|
current
|
36 (12.72)
|
19 (14.18)
|
17 (11.41)
|
0.0006
|
former
|
70 (24.73)
|
46 (34.33)
|
24 (16.11)
|
|
never
|
177 (62.54)
|
69 (51.49)
|
108 (72.48)
|
|
FEV1pred (%)
|
< 83.1
|
222 (78.45)
|
122 (91.04)
|
100 (67.11)
|
< 0.0001
|
≥ 83.1
|
61 (21.55)
|
12 (8.96)
|
49 (32.89)
|
|
FEV1/FVC (%)
|
< 79.77
|
135 (47.70)
|
99 (73.88)
|
36 (24.16)
|
< 0.0001
|
≥ 79.77
|
148 (52.30)
|
35 (26.12)
|
113 (75.84)
|
|
Family history of respiratory disease (%)
|
no
|
264 (93.29)
|
120 (89.55)
|
144 (96.64)
|
0.0322
|
yes
|
19 (6.71)
|
14 (10.45)
|
5 (3.36)
|
|
Respiratory complications (%)
|
no
|
82 (28.98)
|
14 (10.45)
|
68 (45.64)
|
< 0.0001
|
yes
|
201 (71.02)
|
120 (89.55)
|
81 (54.36)
|
|
Hypertension (%)
|
no
|
226 (79.86)
|
95 (70.90)
|
131 (87.92)
|
0.0006
|
yes
|
57 (20.14)
|
39 (29.10)
|
18 (12.08)
|
|
Diabetes (%)
|
no
|
245 (86.57)
|
110 (82.09)
|
135 (90.60)
|
0.0545
|
yes
|
38 (13.43)
|
24 (17.91)
|
14 (9.40)
|
|
Cardiovascular complications (%)
|
no
|
230 (81.27)
|
102 (76.12)
|
128 (85.91)
|
0.0506
|
yes
|
53 (18.73)
|
32 (23.88)
|
21 (14.09)
|
|
Cerebral apoplexy (%)
|
no
|
272 (96.11)
|
125 (93.28)
|
147 (98.66)
|
0.0426
|
yes
|
11 (3.89)
|
9 (6.72)
|
2 (1.34)
|
|
Psychiatric complications (%)
|
no
|
263 (92.93)
|
122 (91.04)
|
141 (94.63)
|
0.3456
|
yes
|
20 (7.07)
|
12 (8.96)
|
8 (5.37)
|
|
Chronic renal complications (%)
|
no
|
255 (90.11)
|
115 (85.82)
|
140 (93.96)
|
0.0366
|
yes
|
28 (9.89)
|
19 (14.18)
|
9 (6.04)
|
|
Funnel chest (%)
|
no
|
215 (75.97)
|
126 (94.03)
|
89 (59.73)
|
< 0.0001
|
yes
|
68 (24.03)
|
8 (5.97)
|
60 (40.27)
|
|
Immune related disease (%)
|
no
|
221 (78.09)
|
117 (87.31)
|
104 (69.80)
|
0.0006
|
yes
|
62 (21.91)
|
17 (12.69)
|
45 (30.20)
|
|
CT signs of emphysema (%)
|
no
|
219 (77.39)
|
91 (67.91)
|
128 (85.91)
|
0.0005
|
yes
|
64 (22.61)
|
43 (32.09)
|
21 (14.09)
|
|
NLR. (median [IQR])
|
2.420 [1.682, 4.003]
|
2.490 [1.955, 3.765]
|
2.283 [1.489, 5.092]
|
0.4178
|
PLR. (median [IQR])
|
118.462 [85.527, 157.621]
|
120.515 [86.490, 159.595]
|
118.227 [81.437, 155.102]
|
0.8109
|
MLR. (median [IQR])
|
0.263 [0.190, 0.398]
|
0.270 [0.212, 0.380]
|
0.258 [0.176, 0.436]
|
0.4346
|
Red blood cell (median [IQR])
|
4.570 [4.200, 4.990]
|
4.540 [4.270, 5.020]
|
4.570 [4.130, 4.920]
|
0.3675
|
Hemoglobin (median [IQR])
|
136.000 [123.000, 147.000]
|
136.000 [122.000, 147.000]
|
136.000 [124.000, 146.000]
|
0.8757
|
Platelet (median [IQR])
|
184.000 [144.000, 228.000]
|
174.500 [146.250, 221.250]
|
192.000 [135.000, 233.000]
|
0.2093
|
White blood cell (median [IQR])
|
6.450 [5.245, 8.125]
|
6.455 [5.217, 7.755]
|
6.450 [5.360, 8.620]
|
0.2223
|
Neutrophil (median [IQR])
|
4.080 [2.970, 5.580]
|
4.200 [3.055, 5.345]
|
3.970 [2.820, 5.870]
|
0.8557
|
Lymphocyte (median [IQR])
|
1.600 [1.215, 2.000]
|
1.525 [1.228, 1.883]
|
1.650 [1.200, 2.120]
|
0.114
|
Monocyte (median [IQR])
|
0.430 [0.320, 0.585]
|
0.430 [0.322, 0.555]
|
0.430 [0.320, 0.660]
|
0.354
|
Eosinophilic (median [IQR])
|
0.120 [0.060, 0.220]
|
0.120 [0.070, 0.238]
|
0.110 [0.050, 0.210]
|
0.1431
|
Basophil (median [IQR])
|
0.020 [0.020, 0.040]
|
0.020 [0.020, 0.040]
|
0.020 [0.020, 0.030]
|
0.5675
|
Albumin (median [IQR])
|
42.000 [38.550, 45.300]
|
40.850 [38.425, 44.600]
|
42.900 [39.400, 45.500]
|
0.0437
|
Glucose (median [IQR])
|
5.120 [4.655, 5.670]
|
5.265 [4.762, 5.885]
|
4.930 [4.600, 5.430]
|
0.0035
|
Triglyceride (median [IQR])
|
1.080 [0.760, 1.480]
|
1.210 [0.825, 1.708]
|
0.980 [0.730, 1.300]
|
0.0019
|
Cholesterol (median [IQR])
|
3.980 [3.430, 4.775]
|
4.305 [3.750, 5.008]
|
3.740 [3.330, 4.460]
|
< 0.0001
|
Fibrinogen (median [IQR])
|
2.930 [2.345, 3.850]
|
3.095 [2.590, 4.058]
|
2.820 [2.250, 3.600]
|
0.0037
|
Data are n (%), or median [IQR]. |
P-value for comparison betweenprogression and stable group.P value less than 0.05 is considered significant. |
Abbreviations:FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; IQR, interquartile range; MLR, monocyte-to-lymphocyte ratio; NLR, Neutrophil to lymphocyte ratio; PLR, Platelet lymphocyte ratio; PRISm, preserved ratio impaired spirometry. |
Table 1 shows the clinical and demographic characteristics of patients in the two cohorts grouped by progression or not. Compared with the progression-free group, patients in the progression group were older, had more former and current smokers, had lower FEV1 pred, lower FEV1/FVC, had a higher rate of family history of respiratory disease, respiratory complications, cerebral apoplexy, chronic kidney disease, and had a higher rate of CT showing emphysema. Laboratory tests also showed significant differences in albumin, glucose, triglycerides, cholesterol, and fibrinogen between the two groups. In addition, during the data collection process, we found that the progression-free group showed a higher rate of funnel chest and immune system disease compared to the progressive group. (P < 0.05 for all, Table 1). It is important to note that the progression-free group included many patients diagnosed with funnel chest, a condition predominantly found in adolescents according to previous studies.[24] Therefore, their clinical data may not be representative of the condition of primarily stable PRISm patients. Including more representative populations is necessary.
In our comparative analysis of baseline data from the training cohort (N = 227) and validation cohort (N = 56), we observed no statistically significant differences in baseline demography and laboratory tests. Therefore, it is reasonable to use these cohorts as training and validation cohort for the model. (Supplementary table 1)
3.2 Multivariate analyses of risk predictors of lung function progression and create the prediction model using the training cohort
We first screened out the factors with significant differences (P < 0.05) between the two cohorts based on the baseline information. Subsequent univariate analysis showed that age 40 ~ 59 (OR = 0.40), BMI low (OR = 0.16), current smoker(OR = 3.22), FEV1 pred ≥ 83.1 (OR = 0.23), FEV1/FVC ≥ 79.77 (OR = 0.10), family history of respiratory disease (OR = 5.19), respiratory complications (OR = 7.30), hypertension (OR = 3.36), chronic renal complications (OR = 2.93), funnel chest (OR = 0.10), immune related disease (OR = 0.27), CT signs of emphysema (OR = 3.53) and glucose high (OR = 2.45) were related to PRISm progression (all P < 0.05); The results of univariate analysis of predictors of PRISm progression are reported in Supplementary table 2; Therefore, we included all the variables with significant baseline differences (all P < 0.05) into the multivariate logistic regression model, and finally selected 7 variables including age 40 ~ 59 (OR = 0.252), BMI low (OR = 0.175), FEV1 pred ≥ 83.1 (OR = 0.175), FEV1/FVC ≥ 79.77 (OR = 0.349), family history of respiratory disease(OR = 5.161), respiratory complications (OR = 6.891), immune related disease(OR = 0.098) to establish the novel nomogram model. The result of the multiple logistic regression of progression for PRISm in the training cohort is reported in Table 2, and the ORs for each risk factor with a 95% CI were presented using a forest plot. (Fig. 2). We can easily draw the conclusion that complicity with respiratory disease and family history of respiratory disease were positively correlated with lung function progression, however, patients aged 40 ~ 59, with lower BMI, higher FEV1 pred, higher FEV1/FVC, and combined immune system-related disease are at lower risk for lung function progression.
Table 2
Multiple logistic regression of progression for PRISm in the training cohort
|
Progression
|
|
Variable
|
OR
|
95%LCI
|
95%UCI
|
P-value
|
Age 40–59
|
0.252
|
0.101
|
0.598
|
0.002282244
|
BMI low
|
0.175
|
0.057
|
0.52
|
0.001864554
|
FEV1pred ≥ 83.1
|
0.175
|
0.058
|
0.49
|
0.001303241
|
FEV1/FVC ≥ 79.77
|
0.349
|
0.151
|
0.807
|
0.013201192
|
Family history of respiratory disease
|
5.161
|
1.224
|
28.169
|
0.035937566
|
Respiratory complications
|
6.891
|
2.6
|
19.148
|
0.000135439
|
Immune related disease
|
0.098
|
0.035
|
0.256
|
0.00000477
|
P value less than 0.05 is considered significant. All variables with meaningful p-values screened by multivariate logistic regression are included in the table. |
Abbreviations: BMI, body mass index; CI, confidence interval; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 s; OR, odds ratio; PRISm, preserved ratio impaired spirometry. |
Each risk predictor (except for family history of respiratory disease, respiratory complications, immune related disease) was evaluated for its prediction performance in the training cohort of PRISm progression. (Supplementary Fig. 1)
3.4 Apparent Performance of nomograms to predict PRISm progression
The estimated progress of the nomogram model to predict the progression of PRISm in the training cohort has a good correlation with the actual progress. Our results show that the calibration curve almost coincides with the Y = X line, indicating that the model is well calibrated. (Fig. 4)
To assess the accuracy of our model, we performed ROC analyses on patients. The AUCs for discriminating between patients with and without lung function progression were 0.890 (95% CI 0.848 to 0.931), and 0.859(95% CI 0.752 to 0.966) in our training and validation cohorts, respectively. (Fig. 5) The results suggest that the AUCs for predicting the risk factors is reliable, which also indicates that the model has good predictive potential. In addition, DCA results show that nomogram models have good clinical application value. The gray curve shows that all patients received the intervention, the black line shows that no patients received the intervention, and the red curve shows the clinical benefit of our model. (Supplementary Fig. 2) CIC showed that the "high risk number" line was relatively close to the "high risk number with event" line, indicating that using this nomogram model to predict PRISm progression has a large clinical net benefit. (Supplementary Fig. 3)