Table 3.
EHR-based observational heritability (ho2) of eGFR, CKD, and albuminuria.
| Number families | ho2 | 95L | 95U | SE | P | Number attempts | Number converged | Number significant | POSA | |
|---|---|---|---|---|---|---|---|---|---|---|
| Estimated GFR | ||||||||||
| Any CKD | 2623 | 0.214 | 0.142 | 0.286 | 0.054 | 4.3E-05 | 200 | 200 | 200 | 1.00 |
| Any CKD White | 919 | 0.197 | 0.131 | 0.321 | 0.080 | 7.1E-03 | 200 | 197 | 152 | 0.77 |
| Any CKD Black | 459 | 0.244 | 0.167 | 0.377 | 0.109 | 1.3E-02 | 200 | 196 | 152 | 0.78 |
| Moderate CKD | 456 | 0.237 | 0.189 | 0.377 | 0.115 | 2.0E-02 | 200 | 139 | 16 | 0.12 |
| Advanced CKD | 131 | 0.455 | 0.369 | 0.656 | 0.186 | 9.6E-03 | 200 | 107 | 10 | 0.09 |
| Albuminuria | ||||||||||
| Any albuminuria | 1122 | 0.236 | 0.180 | 0.366 | 0.439 | 1.8E-02 | 200 | 181 | 53 | 0.29 |
| Severe albuminuria | 417 | 0.490 | 0.364 | 0.864 | 0.313 | 1.5E-02 | 200 | 171 | 73 | 0.43 |
| Any CKD | ||||||||||
| All | 3460 | 0.290 | 0.211 | 0.410 | 0.064 | 4.2E-06 | 200 | 5 | 5 | 1.00 |
| Hispanic | 3136 | 0.251 | 0.185 | 0.296 | 0.091 | 3.1E-05 | 200 | 15 | 15 | 1.00 |
| White | 977 | 0.323 | 0.234 | 0.515 | 0.114 | 7.8E-03 | 200 | 197 | 137 | 0.70 |
| Black | 433 | 0.435 | 0.291 | 0.657 | 0.187 | 1.3E-02 | 200 | 195 | 128 | 0.66 |
| Moderate CKD | ||||||||||
| All | 1529 | 0.618 | 0.418 | 0.822 | 0.089 | 1.1E-08 | 200 | 197 | 197 | 1.00 |
| White | 310 | 0.555 | 0.315 | 0.845 | 0.411 | 9.1E-03 | 200 | 199 | 174 | 0.87 |
| Hispanic | 1024 | 0.513 | 0.298 | 0.781 | 0.251 | 6.3E-05 | 200 | 199 | 197 | 0.99 |
| Black | 174 | 0.777 | 0.544 | 0.988 | 0.351 | 3.4E-02 | 200 | 142 | 87 | 0.61 |
| Advanced CKD | ||||||||||
| All | 537 | 0.761 | 0.512 | 0.964 | 0.086 | 7.0E-07 | 200 | 186 | 186 | 1.00 |
| White | 112 | 0.639 | 0.398 | 0.928 | 0.263 | 3.7E-02 | 200 | 186 | 128 | 0.69 |
| Hispanic | 344 | 0.727 | 0.434 | 0.979 | 0.322 | 2.3E-03 | 200 | 175 | 170 | 0.97 |
| Black | 70 | 0.815 | 0.540 | 0.993 | 0.226 | 1.0E-02 | 200 | 79 | 30 | 0.38 |
The estimates were based on the pedigrees built with RIFTEHR (Relationship Inference From The Electronic Health Record). Estimated GFR was modeled as a quantitative trait, while albuminuria and CKD as dichotomous traits. The numbers of families with available phenotypes is provided, race/ethnicity was determined by self-report for >50% relatives per pedigree. All estimates are adjusted for age, sex, race/ethnicity, and common environment. The SOLARStrap algorithm was run 200 times, subsampling 15-30% of families per run. We used the proportion of significant attempts (POSA) as a quality score for the heritability estimates generated by SOLARStrap as described by Polubriaginof et al. (Cell, 2018). ho2: estimated observational heritability; 95L: lower bound of 95% confidence interval for ho2; 95U: upper bound of 95% confidence interval for ho2; SE: standard error. Race/ethnicity is defined by self-report and determined by the majority of family members.