Table 1 by baseline fluctuation status

	No (N=86)	Yes (N=92)	Overall (N=178)
Baseline age
Mean (SD)	70.4 (6.42)	70.2 (6.55)	70.3 (6.47)
Median [Min, Max]	71.0 [55.0, 83.0]	69.0 [57.0, 89.0]	70.0 [55.0, 89.0]
Missing	15 (17.4%)	8 (8.7%)	23 (12.9%)
Gender
1	62 (72.1%)	74 (80.4%)	136 (76.4%)
2	9 (10.5%)	10 (10.9%)	19 (10.7%)
Missing	15 (17.4%)	8 (8.7%)	23 (12.9%)
Baseline MoCA
Mean (SD)	21.5 (4.71)	18.4 (5.93)	19.8 (5.61)
Median [Min, Max]	21.0 [11.0, 30.0]	19.0 [6.00, 30.0]	21.0 [6.00, 30.0]
Missing	19 (22.1%)	9 (9.8%)	28 (15.7%)
Baseline UPDRS score
Mean (SD)	27.8 (16.0)	30.5 (16.1)	29.3 (16.0)
Median [Min, Max]	26.0 [0, 73.0]	29.0 [2.00, 80.0]	27.5 [0, 80.0]
Missing	19 (22.1%)	11 (12.0%)	30 (16.9%)
Baseline CDR score
Mean (SD)	4.17 (2.50)	5.85 (3.04)	5.09 (2.92)
Median [Min, Max]	3.50 [1.00, 12.0]	4.50 [1.00, 13.0]	4.00 [1.00, 13.0]
Missing	45 (52.3%)	42 (45.7%)	87 (48.9%)
apoe
APOEe4-	44 (51.2%)	45 (48.9%)	89 (50.0%)
APOEe4+	19 (22.1%)	36 (39.1%)	55 (30.9%)
Missing	23 (26.7%)	11 (12.0%)	34 (19.1%)

Cross-sectional relationships at each timepoint

UPDRS scores vs. Fluctuation status

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$updrs_base by dlbc$fluc_base
## t = -1.0284, df = 141.25, p-value = 0.3055
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -7.949400  2.508825
## sample estimates:
##  mean in group No mean in group Yes 
##          27.76119          30.48148

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$updrs_m12 by dlbc$fluc_m12
## t = -0.61146, df = 53.053, p-value = 0.5435
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -10.90161   5.80762
## sample estimates:
##  mean in group No mean in group Yes 
##          28.82143          31.36842

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$updrs_m24 by dlbc$fluc_m24
## t = -2.0433, df = 28.663, p-value = 0.05031
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -24.12083778   0.01726636
## sample estimates:
##  mean in group No mean in group Yes 
##          25.06250          37.11429

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$updrs_m36 by dlbc$fluc_m36
## t = -2.2591, df = 23.991, p-value = 0.03324
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -28.262507  -1.275954
## sample estimates:
##  mean in group No mean in group Yes 
##          24.76923          39.53846

MoCA scores vs. Fluctuation status

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$moca_base by dlbc$fluc_base
## t = 3.5319, df = 147.97, p-value = 0.0005505
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  1.351430 4.784516
## sample estimates:
##  mean in group No mean in group Yes 
##          21.47761          18.40964

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$moca_m12 by dlbc$fluc_m12
## t = 1.5602, df = 52.86, p-value = 0.1247
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -0.6784699  5.4284699
## sample estimates:
##  mean in group No mean in group Yes 
##          20.17857          17.80357

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$moca_m24 by dlbc$fluc_m24
## t = 1.5331, df = 33.354, p-value = 0.1347
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -1.098709  7.828548
## sample estimates:
##  mean in group No mean in group Yes 
##          19.68750          16.32258

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$moca_m36 by dlbc$fluc_m36
## t = 1.0004, df = 20.034, p-value = 0.3291
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -3.186568  9.060694
## sample estimates:
##  mean in group No mean in group Yes 
##          19.84615          16.90909

CDR vs. Fluctuation status

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$cdr_base by dlbc$fluc_base
## t = -2.8911, df = 88.997, p-value = 0.004823
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -2.8333765 -0.5251601
## sample estimates:
##  mean in group No mean in group Yes 
##          4.170732          5.850000

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$cdr_m12 by dlbc$fluc_m12
## t = -1.0775, df = 25.817, p-value = 0.2912
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -3.1249961  0.9759577
## sample estimates:
##  mean in group No mean in group Yes 
##          5.156250          6.230769

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$cdr_m24 by dlbc$fluc_m24
## t = -3.3085, df = 29.073, p-value = 0.002506
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -5.798238 -1.368429
## sample estimates:
##  mean in group No mean in group Yes 
##          3.583333          7.166667

## 
##  Welch Two Sample t-test
## 
## data:  dlbc$cdr_m36 by dlbc$fluc_m36
## t = -2.2259, df = 12.895, p-value = 0.04449
## alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
## 95 percent confidence interval:
##  -7.8853677 -0.1146323
## sample estimates:
##  mean in group No mean in group Yes 
##                 5                 9

Longitudinal relationships by baseline fluctuation status

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 401 rows containing non-finite outside the scale range
## (`stat_smooth()`).

## Warning: Removed 401 rows containing missing values or values outside the scale range
## (`geom_point()`).

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 410 rows containing non-finite outside the scale range
## (`stat_smooth()`).

## Warning: Removed 410 rows containing missing values or values outside the scale range
## (`geom_point()`).

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 513 rows containing non-finite outside the scale range
## (`stat_smooth()`).

## Warning: Removed 513 rows containing missing values or values outside the scale range
## (`geom_point()`).

Measuring consistency

## # A tibble: 10 × 5
## # Groups:   fluc_base, fluc_m12, fluc_m24 [6]
##    fluc_base fluc_m12 fluc_m24 fluc_m36     n
##    <chr>     <chr>    <chr>    <chr>    <int>
##  1 No        No       No       No           1
##  2 No        No       No       Yes          3
##  3 No        No       Yes      No           2
##  4 No        Yes      Yes      No           4
##  5 No        Yes      Yes      Yes          4
##  6 Yes       No       No       No           2
##  7 Yes       No       No       Yes          1
##  8 Yes       Yes      No       No           1
##  9 Yes       Yes      Yes      No           2
## 10 Yes       Yes      Yes      Yes          5

## # A tibble: 35 × 4
##    fluc_base fluc_m12 fluc_m24 fluc_m36
##    <chr>     <chr>    <chr>    <chr>   
##  1 <NA>      Yes      No       No      
##  2 Yes       Yes      Yes      <NA>    
##  3 No        No       No       <NA>    
##  4 Yes       Yes      No       No      
##  5 <NA>      <NA>     <NA>     <NA>    
##  6 No        Yes      No       <NA>    
##  7 Yes       Yes      Yes      No      
##  8 Yes       No       No       No      
##  9 <NA>      Yes      <NA>     <NA>    
## 10 No        Yes      Yes      Yes     
## # ℹ 25 more rows

	Overall (N=203)
Group
Consistent no	17 (8.4%)
Consistent yes	50 (24.6%)
Improvement	15 (7.4%)
Inconsistent	12 (5.9%)
Progression	21 (10.3%)
Missing	88 (43.3%)

	Overall (N=115)
Group
Consistent no	17 (14.8%)
Consistent yes	50 (43.5%)
Improvement	15 (13.0%)
Inconsistent	12 (10.4%)
Progression	21 (18.3%)

Spaghetti plots for people who “progressed”

Spaghetti plots for people who “improved”

Spaghetti plots for consistent yes

Modeling

Cross-sectional models at baseline

	CDR			MoCA			UPDRS
Predictors	Estimates	CI	p	Estimates	CI	p	Estimates	CI	p
(Intercept)	-2.68	-9.03 – 3.68	0.405	28.95	19.31 – 38.59	<0.001	12.02	-16.10 – 40.14	0.399
fluc_baseYes	2.04	0.86 – 3.22	0.001	-2.52	-4.39 – -0.65	0.009	2.74	-2.70 – 8.19	0.321
Baseline age	0.09	-0.00 – 0.18	0.051	-0.10	-0.24 – 0.03	0.141	0.21	-0.19 – 0.61	0.305
Gender: GENDER2	0.78	-0.97 – 2.54	0.378	-1.07	-3.82 – 1.68	0.443	2.00	-5.86 – 9.85	0.616
apoeAPOEe4+	0.15	-1.04 – 1.35	0.798	-1.03	-2.94 – 0.87	0.285	0.06	-5.47 – 5.59	0.983
Observations	87			133			134
R² / R² adjusted	0.174 / 0.134			0.088 / 0.059			0.018 / -0.013

Longitudinal models

	CDR			MoCA			UPDRS
Predictors	Estimates	CI	p	Estimates	CI	p	Estimates	CI	p
(Intercept)	-5.11	-11.92 – 1.69	0.140	32.96	22.26 – 43.65	<0.001	5.32	-24.64 – 35.27	0.727
time	1.43	1.01 – 1.85	<0.001	-2.07	-2.67 – -1.47	<0.001	3.11	1.54 – 4.69	<0.001
fluc_base: Yes	2.85	1.22 – 4.48	0.001	-3.46	-5.96 – -0.96	0.007	1.79	-5.03 – 8.62	0.605
Baseline age	0.10	0.01 – 0.20	0.038	-0.13	-0.28 – 0.02	0.097	0.26	-0.16 – 0.69	0.220
Gender: GENDER2	0.70	-1.20 – 2.60	0.471	-1.74	-4.79 – 1.32	0.264	3.22	-5.20 – 11.65	0.452
apoeAPOEe4+	0.46	-0.81 – 1.74	0.474	-0.92	-3.02 – 1.19	0.392	0.29	-5.57 – 6.15	0.923
time:fluc_baseYes	-0.64	-1.24 – -0.04	0.038	0.82	-0.03 – 1.67	0.058	0.20	-1.98 – 2.37	0.859
Random Effects
σ²	3.26			9.55			66.65
τ₀₀	6.71 _{DLBC_ID}			29.45 _{DLBC_ID}			240.73 _{DLBC_ID}
ICC	0.67			0.76			0.78
N	88 _{DLBC_ID}			133 _{DLBC_ID}			136 _{DLBC_ID}
Observations	195			284			296
Marginal R² / Conditional R²	0.182 / 0.732			0.108 / 0.782			0.046 / 0.793

DLBC Fluctuation Paper 092524

Karl Brown

2024-09-25

Table 1 by baseline fluctuation status

Cross-sectional relationships at each timepoint

UPDRS scores vs. Fluctuation status

MoCA scores vs. Fluctuation status

CDR vs. Fluctuation status

Longitudinal relationships by baseline fluctuation status

Measuring consistency

Spaghetti plots for people who “progressed”

Spaghetti plots for people who “improved”

Spaghetti plots for consistent yes

Modeling

Cross-sectional models at baseline

Longitudinal models