A total of 709 stroke patients were admitted during the study period. After excluding patients who refused to participate (n = 0), had missing data (n = 0), had an impairment of consciousness (n = 3), died during hospitalization (n = 1), or were transferred to another hospital or ward (n = 57), 648 patients were screened. Of the eligible patients, 559 patients who were not independently walking on admission were included in the final analysis (Fig. 1).
Table 1 shows the baseline characteristics of the study population. The mean age of the enrolled patients was 75.5 (13.1) years, and 52.8% were male. The median FIM-Walk score of all patients was 1 [1, 3], and most patients were unable to walk 15 m even with assistance. Psychotropic medications were used by 142 patients (25.4%): benzodiazepines by 27 patients (4.8%), hypnotics by 104 patients (18.6%), antipsychotics by 53 patients (9.5%), and antidepressants by 40 patients (7.2%). The results of the between-group comparison showed that the group of psychotropic drug use had significantly lower FIM-total, FIM-cognitive score, rehabilitation units, and total number of medications, and significantly higher premorbid mRS and CCI than the group of non-use. There were no significant differences in age, sex, stroke type, or BRS.
Table 1
Baseline characteristics of participants and group comparisons by with and without psychotropic drug use
| Total (N = 559) | psychotropic drug use (N = 142) | psychotropic drug non-use (N = 417) | P value |
Age, y | 75.5 (13.1) | 76.7 (12.7) | 75.1 (13.3) | 0.318 |
Sex, male | 295 (52.8) | 80 (56.3) | 215 (51.6) | 0.324 |
Stroke type | | | | 0.128 |
Cerebral infarction | 357 (63.9) | 84 (59.2) | 273 (65.5) | |
Cerebral hemorrhage | 171 (30.6) | 46 (32.4) | 125 (30.0) | |
Subarachnoid hemorrhage | 30 (5.4) | 12 (8.5) | 18 (4.3) | |
Stroke history | 136 (24.3) | 36 (25.4) | 100 (24.0) | 0.742 |
Premorbid mRS | 1 [0, 2] | 1 [0, 3] | 1 [0, 2] | 0.041 |
Onset-admission days | 16 [12, 23] | 17 [12, 22] | 16 [12, 23] | 0.973 |
Paralysis | | | | |
Right / Left / Both | 228 (40.8) /239 (42.8) / 23 (4.1) | 49 (34.5) / 62 (43.7) / 6 (4.2) | 179 (42.9) / 177 (42.4) / 17 (4.1) | 0.072 |
BRS | | | | |
Upper limb / Hand-finger / Lower limb | 5 [2, 6] / 5 [2, 6] /5 [3, 6] | 5 [2, 6] / 5 [2, 6] /5 [3, 6] | 5 [2, 6] / 5 [2, 6] /5 [3, 6] | 0.765 /0.498 /0.209 |
FIM, score | | | | |
-Total | 58 [29, 81] | 48 [29, 71] | 60 [30, 83] | 0.014 |
-Motor | 38 [17, 58] | 34 [18, 51] | 39 [17, 60] | 0.064 |
-Walk | 1 [1, 3] | 1 [1, 2] | 1 [1, 4] | 0.117 |
-Cognitive | 18 [10, 24] | 14 [9, 21] | 19 [11, 25] | < 0.001 |
CCI, score | 3 [1, 4] | 3 [2, 4] | 3 [1, 3] | 0.008 |
Nutritional status | | | | |
MNA-SF | 7 [5, 9] | 7 [5, 9] | 7 [5, 9] | 0.418 |
BMI, kg/m2 | 22.3 [19.7, 24.7] | 21.8 [19.7, 24.9] | 22.3 [19.6, 24.7] | 0.788 |
Energy intake, kcal/kg/day | 26.9 [23.5, 31.0] | 26.9 [23.2, 30.9] | 27.0 [23.7, 31.3] | 0.464 |
Handgrip strength, kg | | | | |
Male | 24.1 [16.2, 31.4] | 26.5 [14.1, 31.1] | 23.7 [17.0, 31.9] | 0.627 |
Female | 11.4 [5.2, 17.0] | 11.0 [5.8, 15.7] | 11.5 [5.1, 17.3] | 0.884 |
Length of stay, days | 91 [61, 135] | 86 [68, 131] | 91 [60, 137] | 0.923 |
Rehabilitation a, units/day | 8.2 [7.0, 8.6] | 7.9 [6.3, 8.5] | 8.3 [7.3, 8.6] | 0.001 |
medication data | | | | |
Number of total drugs | 6 [4, 8] | 7 [5, 9] | 5 [3, 7] | < 0.001 |
Number of psychotropic drugs | 0 [0, 1] | 1 [1, 2] | 0 [0, 0] | < 0.001 |
Benzodiazepine | - | 27 (4.8) | - | - |
Hypnotic | - | 104 (18.6) | - | - |
Antipsychotic | - | 53 (9.5) | - | - |
Antidepressant | - | 40 (7.2) | - | - |
a Rehabilitation therapy (including physical, occupational, and speech and swallowing therapy) performed during hospitalization (1 unit = 20 min). |
BMI, body mass index; BRS, Brunnstrom Recovery Stage; CCI, Charlson’s Comorbidity Index; FIM, Functional Independence Measure; MNA-SF, Mini Nutritional Assessment-Short Form; mRS, modified Rankin Scale. |
Data are expressed as means (standard deviation) for parametric data, while medians and 25th to 75th percentiles (interquartile range (IQR)) were used to describe nonparametric data, and numbers (%) were used to describe categorical data. |
Comparisons between the two groups were made, depending on the type of variable data, using t-tests (two independent variables that were normally distributed), Mann-Whitney U tests (two independent variables that were not normally distributed), and chi-square tests (nominal variables). |
Table 2 shows the comparison of outcomes between psychotropic drug use and no use. On univariate analysis, the psychotropic drug use group had significantly lower FIM walk scores (5 [1, 6] vs. 6 [1, 7], P = 0.030), walking independence (37.3% vs. 54.4%, P < 0.001), and FIM-motor scores (61 [36, 83] vs. 77 [44, 87], P = 0.008) at discharge compared with the psychotropic drug non-use group.
Table 2
Univariate analysis of outcomes with and without psychotropic drug use
| Total (N = 559) | psychotropic drug use (N = 142) | psychotropic drug non-use (N = 417) | P value |
FIM-Walk at discharge, score | 6 [1, 7] | 5 [1, 6] | 6 [1, 7] | 0.030 |
Independence of walk | 281 (50.3) | 53 (37.3) | 228 (54.7) | < 0.001 |
FIM-Motor at discharge, score | 75 [40, 87] | 61 [36, 83] | 77 [44, 87] | 0.008 |
FIM, Functional Independence Measure. |
Data are expressed as medians and 25th to 75th percentiles (interquartile range (IQR)) were used to describe nonparametric data, and numbers (%) were used to describe categorical data. |
Comparisons between the two groups were made, depending on the type of variable data, using Mann-Whitney U tests (two independent variables that were not normally distributed), and chi-square tests (nominal variables). |
Table 3 shows the results of the multivariate analysis. There was no multicollinearity. Logistic regression analysis showed that the number of psychotropic medications at admission was independently associated with walking independence (OR: 0.620, 95% CI: 0.428–0.897, p = 0.011). Multiple linear regression analysis showed that the number of psychotropic medications at admission was not associated with FIM motor at discharge (β = -0.035, p = 0.133).
Table 3
Multivariate regression analysis of outcomes at hospital discharge among inpatients requiring assistance with walking after stroke.
| Independence of walk (FIM-walk ≧ 6) | FIM-Motor at discharge |
| OR (95% CI) | P value | β | B (95% CI) | P value |
Age | 0.922 (0.894, 0.951) | < 0.001 | -0.16 | -0.324 (-0.433, -0.215) | < 0.001 |
Sex (Male) | 0.817 (0.412, 1.622) | 0.564 | -0.05 | -2.645 (-5.421, 0.132) | 0.062 |
Stroke type | | | | | |
Cerebral infarction | 0.558 (0.130, 2.391) | 0.432 | -0.055 | -3.038 (-8.191, 2.114) | 0.247 |
Cerebral hemorrhage | 0.377 (0.086, 1.658) | 0.197 | -0.044 | -2.521 (-7.849, 2.807) | 0.353 |
Subarachnoid hemorrhage | | | | | |
Stroke history | 1.179 (0.600, 2.318) | 0.632 | -0.01 | -0.608 (-3.36, 2.144) | 0.664 |
Premorbid mRS | 0.752 (0.593, 0.953) | 0.018 | -0.112 | -2.077 (-3.057, -1.096) | < 0.001 |
BRS-lower limb | 1.339 (1.052, 1.705) | 0.018 | 0.167 | 2.504 (1.637, 3.371) | < 0.001 |
FIM-Walk on admission | 1.065 (0.827, 1.372) | 0.625 | -0.076 | -1.271 (-2.274, -0.268) | 0.013 |
FIM-Motor on admission (excluded FIM-Walk) | 1.076 (1.045, 1.109) | < 0.001 | 0.418 | 0.544 (0.426, 0.662) | < 0.001 |
FIM-Cognitive on admission | 1.055 (1.003, 1.109) | 0.039 | 0.12 | 0.391 (0.178, 0.603) | < 0.001 |
CCI | 0.785 (0.645, 0.956) | 0.016 | -0.027 | -0.493 (-1.308, 0.321) | 0.235 |
Handgrip strength | 1.042 (1.001, 1.084) | 0.044 | 0.264 | 0.581 (0.421, 0.741) | < 0.001 |
Number of drugs on admission | 1.023 (0.918, 1.140) | 0.683 | -0.002 | -0.018 (-0.44, 0.404) | 0.933 |
Number of psychotropic drugs | 0.620 (0.428, 0.897) | 0.011 | -0.035 | -1.142 (-2.633, 0.349) | 0.133 |
BRS, Brunnstrom Recovery Stage; CCI, Charlson’s Comorbidity Index; FIM, Functional Independence Measure; mRS, modified Rankin Scale. |
Figure 2 shows the results of the logistic regression analysis for each psychotropic drug. Walking independence was significantly associated with hypnotics (OR: 0.331, 95% CI: 0.154–0.708, p = 0.004). There were no statistically significant associations with polypharmacy defined as 5 or more drugs (OR: 1.536, 95% CI: 0.807–2.921, p = 0.191), benzodiazepines (OR: 0.429, 95% CI: 0.112–1.643, p = 0.217), antipsychotics (OR: 0.734, 95% CI: 0.289–1.866, p = 0.517), and antidepressants (OR: 0.475, 95% CI: 0.171–1.381, p = 0.153).