Characteristics of study participants
We analyzed a cohort of 2765 participants aged 60 years and older, using data obtained from NHANES in two phases (2011–2012 and 2013–2014). Table 1 provides a comprehensive overview of the statistical analyses conducted on the overall characteristics of the study population, stratified by quartiles of ALT/AST. The percentages of participants scoring < 5 on the CERAD scale, < 14 on the AF scale, and < 34 on the DSST scale, indicative of cognitive impairment, were found to be 25.21%, 29.55%, and 25.42%, respectively. While several variables, such as education, PIR, smoking status, history of coronary heart disease, history of hypertension, and depressive symptoms, did not show significant differences (p > 0.05), the majority of participants demonstrated statistically significant variations when grouped according to serum ALT/AST quartiles.
Table 1
General characteristics of participants (n = 2765) stratified by ALT/AST in the NAHENS 2011–2014.
| Total | Q1 | Q2 | Q3 | Q4 | P-value |
N | 2765 | 684 | 688 | 605 | 788 | |
Gender | | | | | | < 0.001 |
Male | 1356 (49.04%) | 286 (41.81%) | 297 (43.17%) | 298 (49.26%) | 475 (60.28%) | |
Female | 1409 (50.96%) | 398 (58.19%) | 391 (56.83%) | 307 (50.74%) | 313 (39.72%) | |
Age (years) | | | | | | < 0.001 |
60–69 | 1500 (54.25%) | 289 (42.25%) | 331 (48.11%) | 361 (59.67%) | 519 (65.86%) | |
70–79 | 809 (29.26%) | 198 (28.95%) | 232 (33.72%) | 171 (28.26%) | 208 (26.40%) | |
≥80 | 456 (16.49%) | 197 (28.80%) | 125 (18.17%) | 73 (12.07%) | 61 (7.74%) | |
Race | | | | | | < 0.001 |
Mexican American/other Hispanic | 521 (18.84%) | 86 (12.57%) | 112 (16.28%) | 126 (20.83%) | 197 (25.00%) | |
Non-Hispanic White | 1349 (48.79%) | 364 (53.22%) | 348 (50.58%) | 289 (47.77%) | 348 (44.16%) | |
Non-Hispanic Black | 630 (22.78%) | 177 (25.88%) | 160 (23.26%) | 129 (21.32%) | 164 (20.81%) | |
Non-Hispanic Asian | 227 (8.21%) | 45 (6.58%) | 57 (8.28%) | 54 (8.93%) | 71 (9.01%) | |
Other Race | 38 (1.37%) | 12 (1.75%) | 11 (1.60%) | 7 (1.16%) | 8 (1.02%) | |
Education | | | | | | 0.933 |
Less than high school | 689 (24.92%) | 174 (25.44%) | 173 (25.15%) | 150 (24.79%) | 192 (24.37%) | |
High school | 657 (23.76%) | 172 (25.15%) | 161 (23.40%) | 137 (22.64%) | 187 (23.73%) | |
Above high school | 1417 (51.25%) | 337 (49.27%) | 354 (51.45%) | 318 (52.56%) | 408 (51.78%) | |
Marital status | | | | | | < 0.001 |
Married/living with partner | 1601 (57.90%) | 334 (48.83%) | 381 (55.38%) | 363 (60.00%) | 523 (66.37%) | |
Widowed/divorced/separated | 1004 (36.31%) | 309 (45.18%) | 280 (40.70%) | 204 (33.72%) | 211 (26.78%) | |
Never married | 158 (5.71%) | 41 (5.99%) | 27 (3.92%) | 37 (6.12%) | 53 (6.73%) | |
Poverty-income ratio | | | | | | 0.769 |
< 1 | 423 (15.30%) | 118 (17.25%) | 105 (15.26%) | 89 (14.71%) | 111 (14.09%) | |
> 1 | 2114 (76.46%) | 513 (75.00%) | 524 (76.16%) | 467 (77.19%) | 610 (77.41%) | |
Not recorded | 228 (8.25%) | 53 (7.75%) | 59 (8.58%) | 49 (8.10%) | 67 (8.50%) | |
Alcohol (12 alcohol drinks per year) | | | | | | 0.002 |
Yes | 1866 (68.58%) | 433 (64.53%) | 439 (64.75%) | 426 (70.88%) | 568 (73.67%) | |
No | 852 (31.31%) | 237 (35.32%) | 238 (35.10%) | 174 (28.95%) | 203 (26.33%) | |
Smoked | | | | | | 0.102 |
Yes | 1400 (50.63%) | 352 (51.46%) | 329 (47.82%) | 318 (52.56%) | 401 (50.89%) | |
No | 1363 (49.29%) | 332 (48.54%) | 359 (52.18%) | 285 (47.11%) | 387 (49.11%) | |
Vigorous work activity | | | | | | 0.027 |
Yes | 301 (10.89%) | 59 (8.63%) | 63 (9.16%) | 78 (12.89%) | 101 (12.82%) | |
No | 2463 (89.08%) | 624 (91.23%) | 625 (90.84%) | 527 (87.11%) | 687 (87.18%) | |
BMI (kg/m²) | | | | | | < 0.001 |
< 25 | 734 (26.55%) | 257 (37.57%) | 190 (27.62%) | 153 (25.29%) | 134 (17.01%) | |
25–30 | 980 (35.44%) | 221 (32.31%) | 257 (37.35%) | 204 (33.72%) | 298 (37.82%) | |
> 30 | 1011 (36.56%) | 193 (28.22%) | 230 (33.43%) | 240 (39.67%) | 348 (44.16%) | |
Table 1
| Total | Q1 | Q2 | Q3 | Q4 | P-value |
N | 2765 | 684 | 688 | 605 | 788 | |
Coronary heart disease | | | | | | 0.394 |
Yes | 256 (9.26%) | 69 (10.09%) | 71 (10.32%) | 51 (8.43%) | 65 (8.25%) | |
No | 2494 (90.20%) | 609 (89.04%) | 614 (89.24%) | 553 (91.40%) | 718 (91.12%) | |
Not recorded | 15 (0.54%) | 6 (0.88%) | 3 (0.44%) | 1 (0.17%) | 5 (0.63%) | |
Stroke | | | | | | < 0.001 |
Yes | 192 (6.94%) | 73 (10.67%) | 44 (6.40%) | 30 (4.96%) | 45 (5.71%) | |
No | 2569 (92.91%) | 609 (89.04%) | 643 (93.46%) | 575 (95.04%) | 742 (94.16%) | |
Diabetes | | | | | | < 0.001 |
Yes | 633 (22.89%) | 128 (18.71%) | 138 (20.06%) | 145 (23.97%) | 222 (28.17%) | |
No | 2005 (72.51%) | 532 (77.78%) | 527 (76.60%) | 423 (69.92%) | 523 (66.37%) | |
Not recorded | 125 (4.52%) | 23 (3.36%) | 22 (3.20%) | 37 (6.12%) | 43 (5.46%) | |
Hypertension | | | | | | 0.872 |
Yes | 1718 (62.13%) | 427 (62.43%) | 427 (62.06%) | 373 (61.65%) | 491 (62.31%) | |
No | 1042 (37.69%) | 255 (37.28%) | 259 (37.65%) | 231 (38.18%) | 297 (37.69%) | |
High cholesterol | | | | | | 0.016 |
Yes | 1548 (55.99%) | 355 (51.90%) | 382 (55.52%) | 332 (54.88%) | 479 (60.79%) | |
No | 1198 (43.33%) | 324 (47.37%) | 302 (43.90%) | 271 (44.79%) | 301 (38.20%) | |
Not recorded | 19 (0.69%) | 5 (0.73%) | 4 (0.58%) | 2 (0.33%) | 8 (1.02%) | |
Depressive | | | | | | 0.124 |
Yes | 2520 (91.14%) | 623 (91.08%) | 640 (93.02%) | 552 (91.24%) | 705 (89.47%) | |
No | 245 (8.86%) | 61 (8.92%) | 48 (6.98%) | 53 (8.76%) | 83 (10.53%) | |
ALT (U/L) | 22.22 ± 12.78 | 16.12 ± 6.74 | 19.00 ± 5.86 | 22.73 ± 10.22 | 29.92 ± 17.98 | < 0.001 |
AST (U/L) | 25.24 ± 11.42 | 26.37 ± 12.84 | 24.14 ± 7.49 | 25.01 ± 11.42 | 25.38 ± 12.80 | 0.004 |
CERAD (< 5) | 697 (25.21%) | 213 (31.14%) | 178 (25.87%) | 151 (24.96%) | 155 (19.67%) | < 0.001 |
AF (< 14) | 817 (29.55%) | 240 (35.09%) | 213 (30.96%) | 170 (28.10%) | 194 (24.62%) | < 0.001 |
DSST (< 34) | 703 (25.42%) | 222 (32.46%) | 178 (25.87%) | 134 (22.15%) | 169 (21.45%) | < 0.001 |
Mean +/- SD for: ALT (U/L), AST (U/L). P value was calculated by one-way ANOVA. |
% for: gender, race, education, age (years), BMI (kg/m²), poverty-income ratio, vigorous work activity, alcohol, smoked, coronary heart disease, history of hypertension, history of high cholesterol, history of diabetes, stroke, depressive. P value was calculated by weighted chi-square test. |
In the analysis of serum ALT/AST distribution, it was observed that individuals with higher levels of serum ALT/AST were predominantly male, aged 60–69 years, of Mexican American/other Hispanic descent, Non-Hispanic Asian, married/living with a partner, alcohol consumers, possessing a BMI > 30 kg/m2, and with a medical history comprising diabetes. In contrast, individuals with lower levels of serum ALT/AST were identified as female, 80 years and older, Non-Hispanic White, Non-Hispanic Black, Other Race, widowed/divorced/separated, non-alcohol consumers, with a BMI < 25 kg/m2, and without a medical history of diabetes.
Association between serum ALT/AST and cognitive impairment
Table 2 depicts the correlation between serum ALT/AST levels and cognitive impairment as assessed by the CERAD, AF, and DSST tests. Multivariate logistic regression models were utilized, incorporating three distinct models: model 1 (non-adjusted), model 2 (minor adjusted), and model 3 (fully adjusted).
Table 2
Associations between ALT/AST and cognitive impairment (CERAD < 5, AF < 14, and DSST < 34) (n = 2765), NHANES 2011–2014.
Exposure | Model 1 OR (95% CI) | Model 2 OR (95% CI) | Model 3 OR (95% CI) |
CERAD < 5 | | | |
ALT/AST | 0.36 (0.24, 0.54) < 0.0001 | 0.49 (0.32, 0.75) 0.0010 | 0.53 (0.34, 0.83) 0.0056 |
Quartile of ALT/AST | | | |
Q1 | 1 | 1 | 1 |
Q2 | 0.77 (0.61, 0.98) 0.0308 | 0.87 (0.68, 1.12) 0.2930 | 0.90 (0.70, 1.17) 0.4399 |
Q3 | 0.74 (0.58, 0.94) 0.0140 | 0.90 (0.69, 1.18) 0.4537 | 0.97 (0.74, 1.28) 0.8485 |
Q4 | 0.54 (0.43, 0.69) < 0.0001 | 0.65 (0.50, 0.84) 0.0011 | 0.67 (0.50, 0.88) 0.0043 |
P for trend | < 0.001 | 0.004 | 0.012 |
AF < 14 | | | |
ALT/AST | 0.47 (0.33, 0.68) < 0.0001 | 0.67 (0.46, 0.98) 0.0416 | 0.77 (0.51, 1.15) 0.1937 |
Quartile of ALT/AST | | | |
Q1 | 1 | 1 | 1 |
Q2 | 0.83 (0.66, 1.04) 0.1042 | 0.89 (0.71, 1.13) 0.3468 | 0.96 (0.74, 1.23) 0.7201 |
Q3 | 0.72 (0.57, 0.92) 0.0073 | 0.85 (0.67, 1.10) 0.2191 | 0.92 (0.70, 1.20) 0.5307 |
Q4 | 0.60 (0.48, 0.76) < 0.0001 | 0.74 (0.58, 0.94) 0.0140 | 0.80 (0.61, 1.04) 0.0900 |
P for trend | < 0.001 | 0.029 | 0.106 |
DSST < 34 | | | |
ALT/AST | 0.39 (0.26, 0.58) < 0.0001 | 0.44 (0.28, 0.68) 0.0002 | 0.46 (0.28, 0.77) 0.0027 |
Quartile of ALT/AST | | | |
Q1 | 1 | 1 | 1 |
Q2 | 0.73 (0.57, 0.92) 0.0074 | 0.76 (0.59, 0.98) 0.0378 | 0.79 (0.59, 1.05) 0.1086 |
Q3 | 0.59 (0.46, 0.76) < 0.0001 | 0.65 (0.49, 0.85) 0.0020 | 0.69 (0.50, 0.94) 0.0209 |
Q4 | 0.57 (0.45, 0.72) < 0.0001 | 0.59 (0.45, 0.77) 0.0001 | 0.60 (0.44, 0.82) 0.0012 |
P for trend | < 0.001 | < 0.001 | 0.004 |
Model 1: Non-adjusted model; Model 2: adjust for: gender, age, race; Model 3: adjust for: gender, age, race, education, high blood pressure, diabetes, had at least 12 alcohol drinks lifetime, smoked at least 100 cigarettes in life, BMI (kg/m²). |
(1) Analysis using the CERAD test revealed statistically significant differences in each model when examining the association between serum ALT/AST and the reduced risk of cognitive impairment. In the fully adjusted model, the odds ratio (OR) was determined to be 0.53 (95% confidence interval [CI]: 0.34–0.83). Specifically, participants in the fourth quartile (Q4) of serum ALT/AST exhibited an OR of 0.67 (95% CI: 0.50–0.88), indicating a significant association with cognitive impairment. Additionally, all p-values for trend were calculated to be 0.012.
(2) Findings from the AF test revealed that, in the non-adjusted model, the OR with a 95% CI was 0.47 (0.33, 0.68). Following adjustment for age, gender, and race, a statistically significant association was observed between serum ALT/AST levels and a reduced risk of cognitive impairment (OR = 0.67, 95% CI: 0.46–0.98). However, in the fully adjusted model, no significant differences were observed between serum ALT/AST and cognitive impairment.
(3) Results from the DSST test indicated that, in the non-adjusted model, the OR with a 95% CI was 0.39 (0.26, 0.58). Upon adjusting for age, gender, and race, a statistically significant association was observed between serum ALT/AST levels and a reduced risk of cognitive impairment (OR = 0.44, 95% CI: 0.28–0.68). Furthermore, in the fully adjusted model, a statistically significant association was identified between serum ALT/AST and cognitive impairment [OR (95%CI): 0.46(0.28, 0.77)], as well as for participants in Q3 (OR = 0.69, 95% CI: 0.50–0.94) and in Q4 (OR = 0.60, 95% CI: 0.44–0.82).
Identification of non-linear relationship
The GAM model was employed to explore the potential non-linear relationship between serum ALT/AST levels and cognitive impairment, as illustrated in Fig. 2. After extensive adjustments, the data strongly supported a non-linear correlation between serum ALT/AST and cognitive impairment.
Utilizing a weighted two-piecewise linear regression model with a recursive algorithm, the analysis identified an inflection point of 1.29 based on the CERAD test outcomes (see Table 3). Below or equal to 1.29, the odds ratio (OR) and its corresponding 95% confidence interval (CI) were 0.39 (0.24, 0.65) (p = 0.0003). Conversely, above 1.29, the OR and 95% CI were 2.61 (0.83, 8.17).
Table 3
Nonlinearity addressing by weighted two-piecewise linear model based on CERAD, AF, and DSST tests.
Outcome: | CERAD log2 transform OR (95% CI) | AF log2 transform OR (95% CI) | DSST log2 transform OR (95% CI) |
Fitting by weighted linear regression model | 0.53 (0.34, 0.83) 0.0056 | 0.77 (0.51, 1.15) 0.1937 | 0.46 (0.28, 0.77) 0.0027 |
Fitting by weighted two-piecewise linear regression model | | | |
Inflection point | 1.29 | 0.63 | 0.87 |
< Inflection point | 0.39 (0.24, 0.65) 0.0003 | 0.01 (0.00,0.06) < 0.0001 | 0.14 (0.05, 0.39) 0.0002 |
> Inflection point | 2.61 (0.83, 8.17) 0.0991 | 1.14 (0.74, 1.75) 0.5616 | 1.00 (0.48, 2.07) 0.9976 |
Log likelihood ratio test | 0.019 | < 0.001 | 0.009 |
The association adjusted for: gender, age, race, education, marital status, poverty-income ratio (PIR), high blood pressure, high cholesterol, history of coronary heart disease, history of stroke, diabetes, had at least 12 alcohol drinks lifetime, smoked at least 100 cigarettes in life, vigorous activity over past 30 days, BMI (kg/m²). |
Similarly employing the weighted two-piecewise linear regression model and recursive algorithm, the results revealed an inflection point of 0.63 derived from the AF test results (refer to Table 3). Below the inflection point, the OR with a 95% CI was 0.01 (0.00, 0.06) (p < 0.0001). In contrast, above 0.63, the OR and 95% CI were 1.14 (0.74, 1.75).
Additionally, the weighted two-piecewise linear regression model and recursive algorithm unveiled a critical point of 0.87 based on the DSST test outcomes (as shown in Table 3). Below 0.87, the OR and 95% CI were 0.14 (0.05, 0.39) (p = 0.0002). On the other hand, above 0.87, the OR and 95% CI were 1.00 (0.48, 2.07).
Subgroup analyses outcomes
Subgroup analysis and interaction results are summarized in Table 4.
Table 4
Subgroup analysis of all variables and interaction tests.
ALT/AST | | CERAD | | AF | | DSST | |
| N | ꞵ(95%CI) | p for interaction | ꞵ(95%CI) | p for interaction | ꞵ(95%CI) | p for interaction |
Sex | | | 0.5514 | | 0.2488 | | 0.2875 |
Male | 1356 | 0.38 (0.21, 0.68) 0.0012 | | 0.49 (0.26, 0.95) 0.0340 | | 0.57 (0.32, 1.02) 0.0570 | |
Female | 1409 | 0.90 (0.44, 1.84) 0.7753 | | 0.56 (0.25, 1.26) 0.1625 | | 1.11 (0.64, 1.94) 0.7106 | |
Age (years) | | | 0.8915 | | 0.0065 | | 0.6778 |
60–69 | 1500 | 0.49 (0.25, 0.94) 0.0314 | | 0.53 (0.27, 1.06) 0.0724 | | 1.26 (0.76, 2.10) 0.3746 | |
70–79 | 809 | 0.47 (0.21, 1.08) 0.0770 | | 0.63 (0.24, 1.62) 0.3338 | | 0.62 (0.27, 1.41) 0.2550 | |
≥80 | 456 | 0.53 (0.19, 1.48) 0.2225 | | 0.30 (0.09, 1.06) 0.0611 | | 0.21 (0.07, 0.62) 0.0051 | |
Race | | | 0.0091 | | 0.0004 | | 0.2151 |
Mexican American/other Hispanic | 522 | 1.00 (1.00, 1.00) 0.9501 | | 1.00 (1.00, 1.00) 0.3441 | | 1.00 (1.00, 1.00) 0.3669 | |
Non-Hispanic White | 521 | 0.44 (0.17, 1.13) 0.0876 | | 0.45 (0.16, 1.24) 0.1215 | | 0.34 (0.13, 0.91) 0.0316 | |
Non-Hispanic Black | 1349 | 0.50 (0.25, 1.00) 0.0494 | | 0.30 (0.12, 0.74) 0.0094 | | 0.39 (0.19, 0.80) 0.0105 | |
Non-Hispanic Asian | 630 | 0.94 (0.40, 2.22) 0.8899 | | 0.79 (0.33, 1.87) 0.5879 | | 1.51 (0.72, 3.13) 0.2734 | |
Other Race | 227 | 0.39 (0.05, 2.92) 0.3600 | | 0.37 (0.03, 4.84) 0.4487 | | 2.12 (0.67, 6.73) 0.2028 | |
Education | | | 0.048 | | 0.4084 | | 0.1291 |
Less than high school | 689 | 0.74 (0.34, 1.64) 0.4621 | | 0.58 (0.26, 1.28) 0.1759 | | 0.48 (0.22, 1.03) 0.0586 | |
High school | 657 | 0.59 (0.24, 1.42) 0.2381 | | 0.37 (0.15, 0.93) 0.0344 | | 1.04 (0.47, 2.28) 0.9208 | |
Above high school | 1417 | 0.34 (0.16, 0.71) 0.0044 | | 0.47 (0.17, 1.28) 0.1390 | | 0.77 (0.40, 1.47) 0.4280 | |
Marital status | | | 0.4958 | | 0.2334 | | 0.4764 |
Married/living with partner | 1601 | 0.42 (0.23, 0.77) 0.0049 | | 0.44 (0.22, 0.87) 0.0189 | | 0.70 (0.41, 1.21) 0.2046 | |
Widowed/divorced/separated | 1004 | 0.94 (0.46, 1.95) 0.8761 | | 0.37 (0.16, 0.89) 0.0260 | | 0.84 (0.42, 1.68) 0.6264 | |
Never married | 158 | 0.07 (0.01, 0.72) 0.0252 | | 1.53 (0.23, 10.09) 0.6584 | | 1.23 (0.21, 7.18) 0.8209 | |
Poverty-income ratio | | | 0.398 | | 0.5205 | | 0.7283 |
< 1 | 423 | 0.13 (0.04, 0.47) 0.0019 | | 0.32 (0.09, 1.12) 0.0742 | | 0.35 (0.11, 1.09) 0.0702 | |
> 1 | 2114 | 0.60 (0.36, 1.00) 0.0516 | | 0.45 (0.25, 0.82) 0.0089 | | 0.87 (0.55, 1.37) 0.5400 | |
Not recorded | 228 | 0.66 (0.10, 4.20) 0.6605 | | 0.25 (0.03, 2.14) 0.2054 | | 1.04 (0.21, 5.20) 0.9615 | |
Alcohol (12 alcohol drinks per year) | | | 0.0021 | | 0.0034 | | 0.0429 |
Yes | 1866 | 0.34 (0.19, 0.60) 0.0002 | | 0.35 (0.18, 0.68) 0.0017 | | 0.52 (0.30, 0.89) 0.0166 | |
No | 852 | 1.23 (0.59, 2.54) 0.5840 | | 0.67 (0.30, 1.51) 0.3372 | | 1.40 (0.72, 2.71) 0.3207 | |
Smoked | | | 0.2411 | | 0.5682 | | 0.1663 |
Yes | 1400 | 0.51 (0.27, 0.96) 0.0374 | | 0.52 (0.27, 0.99) 0.0453 | | 0.62 (0.36, 1.08) 0.0907 | |
No | 1363 | 0.52 (0.27, 1.01) 0.0519 | | 0.32 (0.14, 0.72) 0.0064 | | 0.93 (0.49, 1.75) 0.8205 | |
Table 4
ALT/AST | | CERAD | | AF | | DSST | |
| N | ꞵ(95%CI) | p for interaction | ꞵ(95%CI) | p for interaction | ꞵ(95%CI) | p for interaction |
Vigorous work activity | | | 0.38 | | 0.5405 | | 0.0076 |
Yes | 301 | 0.39 (0.08, 1.95) 0.2497 | | 1.15 (0.14, 9.30) 0.8968 | | 0.20 (0.04, 1.10) 0.0636 | |
No | 2463 | 0.54 (0.34, 0.87) 0.0105 | | 0.40 (0.23, 0.67) 0.0006 | | 0.83 (0.55, 1.26) 0.3862 | |
BMI (kg/m²) | | | 0.0578 | | 0.5142 | | 0.9602 |
< 25 | 734 | 0.23 (0.08, 0.66) 0.0062 | | 0.61 (0.20, 1.84) 0.3788 | | 0.59 (0.25, 1.40) 0.2349 | |
25–30 | 980 | 0.50 (0.24, 1.05) 0.0663 | | 0.53 (0.22, 1.26) 0.1521 | | 0.74 (0.36, 1.52) 0.4097 | |
> 30 | 1011 | 0.80 (0.38, 1.68) 0.5536 | | 0.39 (0.17, 0.91) 0.0291 | | 1.04 (0.54, 2.01) 0.9012 | |
Coronary heart disease | | | 0.2869 | | 0.6308 | | 0.5802 |
Yes | 256 | 0.17 (0.03, 0.89) 0.0353 | | 1.85 (0.33, 10.30) 0.4846 | | 0.64 (0.15, 2.73) 0.5431 | |
No | 2494 | 0.59 (0.37, 0.95) 0.0296 | | 0.42 (0.24, 0.72) 0.0015 | | 0.77 (0.50, 1.18) 0.2346 | |
Stroke | | | 0.8173 | | 0.381 | | 0.8431 |
Yes | 192 | 0.47 (0.09, 2.44) 0.3670 | | 0.32 (0.07, 1.54) 0.1546 | | 0.64 (0.13, 3.20) 0.5831 | |
No | 2569 | 0.56 (0.35, 0.89) 0.0153 | | 0.48 (0.28, 0.82) 0.0069 | | 0.75 (0.49, 1.14) 0.1763 | |
Diabetes | | | 0.48 | | 0.1564 | | 0.2367 |
Yes | 633 | 0.72 (0.30, 1.69) 0.4468 | | 0.72 (0.29, 1.79) 0.4774 | | 0.95 (0.42, 2.16) 0.9073 | |
No | 2005 | 0.47 (0.27, 0.82) 0.0076 | | 0.33 (0.18, 0.63) 0.0007 | | 0.65 (0.40, 1.06) 0.0849 | |
Not recorded | 125 | 0.16 (0.01, 2.34) 0.1813 | | 4.88 (0.12, 204.24) 0.4056 | | 7.08 (0.51, 97.45) 0.1435 | |
Hypertension | | | 0.4309 | | 0.2393 | | 0.0097 |
Yes | 1718 | 0.47 (0.27, 0.82) 0.0078 | | 0.38 (0.21, 0.71) 0.0021 | | 0.64 (0.39, 1.07) 0.0898 | |
No | 1042 | 0.61 (0.28, 1.33) 0.2116 | | 0.80 (0.34, 1.89) 0.6156 | | 1.02 (0.51, 2.04) 0.9613 | |
High cholesterol | | | 0.1005 | | 0.0224 | | 0.0296 |
Yes | 1548 | 0.37 (0.20, 0.68) 0.0013 | | 0.53 (0.28, 1.03) 0.0593 | | 0.83 (0.48, 1.44) 0.5081 | |
No | 1198 | 0.84 (0.42, 1.68) 0.6208 | | 0.42 (0.19, 0.93) 0.0333 | | 0.72 (0.39, 1.34) 0.2994 | |
Not recorded | 19 | 0.00 (0.00, Inf) 1.0000 | | Inf (0.00, Inf) 0.9999 | | Inf (0.00, Inf) 0.9999 | |
Depressive | | | 0.9137 | | 0.1626 | | 0.6887 |
Yes | 2520 | 0.49 (0.30, 0.78) 0.0031 | | 0.45 (0.26, 0.78) 0.0039 | | 0.82 (0.54, 1.24) 0.3474 | |
No | 245 | 1.16 (0.22, 6.06) 0.8592 | | 0.59 (0.11, 3.06) 0.5329 | | 0.24 (0.05, 1.14) 0.0736 | |
Each continuous variable was divided into three groups according to its value. Each subgroup analysis adjusted for all the founders except the subgroup factor itself. Potential interactions between ALT/AST and mean cognitive impairment were tested by a likelihood ratio test. |
(1) Analysis of the data revealed significant interaction effects for race, education, and alcohol consumption in relation to serum ALT/AST levels and cognitive impairment as assessed by the CERAD test (interaction p-value < 0.05). Further subgroup analysis confirmed that participants aged 60–69 years, males, non-Hispanic Black individuals, those with education beyond high school, married or cohabiting, never married, with a poverty-income ratio < 1, alcohol consumers, smokers, with a BMI < 25, engaging in less vigorous work, and with no history of stroke, diabetes, coronary heart disease, hypertension, high cholesterol, or depression, were at a reduced risk of cognitive impairment (p < 0.05).
(2) Significant interaction effects were observed between serum ALT/AST levels and cognitive impairment as evaluated by the AF test, with age, race, alcohol consumption, and high cholesterol showing significant interactions (interaction p-value < 0.05). Subgroup analysis revealed a significantly increased risk of cognitive impairment among participants with education beyond high school, non-Hispanic Black individuals, married or cohabiting, widowed/divorced/separated, PIR > 1, alcohol consumers, non-smokers, engaging in less vigorous work, BMI > 30, and with a history of no stroke, diabetes, coronary heart disease, hypertension, high cholesterol, or depression (p < 0.05).
(3) Interaction effects based on the DSST test between serum ALT/AST levels and cognitive impairment indicated significant associations with alcohol consumption, vigorous work, history of hypertension, and high cholesterol (interaction p-value < 0.05). Subgroup analysis suggested that participants aged 80 years and older, non-Hispanic Black and White individuals, PIR < 1, and alcohol consumers had significantly reduced risks of cognitive impairment (all p < 0.05).