Table 3.
Multilevel model of associations between 5-HTTLPR and BDNF genotypes and children’s cortisol activity
| Fixed effect | β | SE | T | P | |
|---|---|---|---|---|---|
| Cortisol intercept | .466 | .024 | 19.297 | .000 | |
| Time of visit | −.081 | .050 | −1.622 | .125 | |
| Age | −.014 | .008 | −1.638 | .103 | |
| Gender | −.078 | .051 | −1.541 | .125 | |
| 5-HTTLPR | .004 | .071 | .063 | .951 | |
| BDNF | .087 | .066 | 1.326 | .187 | |
| 5-HTTLPR×BDNF | −.301 | .113 | −2.656 | .009 | |
| Cortisol linear slope | −.087 | .016 | −5.580 | .000 | |
| Time of visit | .038 | .035 | 1.085 | .280 | |
| Age | .007 | .005 | 1.365 | .175 | |
| Gender | .040 | .033 | 1.210 | .229 | |
| 5-HTTLPR | −.140 | .043 | −3.259 | .002 | |
| BDNF | −.025 | .038 | −.642 | .521 | |
| 5-HTTLPR×BDNF | .272 | .102 | 2.670 | .009 | |
| Cortisol quadratic curvature | .030 | .004 | 8.630 | .000 | |
| Time of visit | −.010 | .007 | −1.416 | .159 | |
| Age | −.001 | .001 | −.838 | .404 | |
| Gender | −.010 | .007 | −1.332 | .185 | |
| 5-HTTLPR | .038 | .009 | 4.226 | .000 | |
| BDNF | .006 | .009 | .649 | .517 | |
| 5-HTTLPR×BDNF | −.060 | .023 | −2.641 | .010 | |
| Random effect of unconditional growth curve |
Variance Component |
SD | χ2 | P | |
| Level 1 intercept | .076 | .275 | 985.294 | .000 | |
| Linear slope | .023 | .152 | 377.517 | .000 | |
| Quadratic curvature | .001 | .034 | 360.028 | .000 | |
Note. N = 144; for t tests, df = 137. All fixed effects are with robust standard errors. All Level 1 predictors are uncentered; Level 2 variables are grand mean centered. SE = standard error; SD = standard deviation; Time of visit: 1000h = 0 and 1400h = 1; Gender: Males = 0 and Females = 1; 5-HTTLPR = serotonin transporter promotor region polymorphism: s/l and l/l = 0 and s/s = 1; BDNF = brain-derived neurotrophic factor Val66Met polymorphism: Val/Val = 0 and Val/Met and Met/Met = 1.