Postdoctoral NIH T32 programs aim to provide select trainees with experience to enable successful career and research development. In a previous posting, we presented NIH career development, or K award, and R01 equivalent research grant outcomes of participants in post-doctoral T32 programs. We showed that men with MD-PhD degrees had the highest rates for R01 application submission and award receipt (as PI). Women comprised a greater proportion of trainees over time. Serving as PI on a K-awards predicted a substantially higher rate of later support as PI on R01 awards. Here we present further analyses focusing on the association of under-represented minority (URM) status with subsequent K or R01 awards.
These analyses are based on the outcomes of 33,977 T32 post-doctoral participants who matriculated between 1995 and 2009, for whom we had data on age, gender, and URM status, and who had received MD, MD-PhD, or PhD degrees. URM status was known in over 95% of matriculants, and was defined as self-identification as Black, Hispanic, American Indian, or Pacific Islander or participation in an NIH diversity program. We followed participants for designation as principal investigator (PI) on subsequent K and R01-equivalent awards (which we’ll refer to as R01 going forward) through 2017. Table 1 shows race and ethnicity according to gender; Whites and Asians were most common.
Table 1: T32 Participant Race and Ethnicity According to Gender
|Race or Ethnicity||Men (%)||Women (%)|
|Total N (%)||17810 (52.4)||16167 (47.6)|
|White||12125 (68.1)||10754 (66.5)|
|Asian||3737 (21.0)||3256 (20.1)|
|Mixed||831 (4.7)||839 (5.2)|
|Black||670 (3.8)||941 (5.8)|
|Hispanic||341 (1.9)||260 (1.6)|
|Native American||54 (0.3)||56 (0.3)|
|Pacific Islander||43 (0.2)||48 (0.3)|
Table 2 shows characteristics and high-level outcomes according to matriculation cohort. Over time, a greater proportion of matriculants were women and, to a lesser extent, from URM groups.
Table 2: T32 Participant Characteristics According to Matriculation Cohort
|Total N (%)||10849 (31.9)||11306 (33.3)||11822 (34.8)|
|Age (years) at Matriculation||30 or less||3387 (31.2)||3852 (34.1)||4200 (35.5)|
|31 to 35||4822 (44.4)||4977 (44.0)||5377 (45.5)|
|36 or more||2640 (24.3)||2477 (21.9)||2245 (19.0)|
|Degree||MD||3250 (30.0)||3953 (35.0)||4080 (34.5)|
|MD-PhD||1240 (11.4)||1191 (10.5)||1095 (9.3)|
|PhD||6359 (58.6)||6162 (54.5)||6647 (56.2)|
|Gender||Men||6034 (55.6)||6010 (53.2)||5766 (48.8)|
|Women||4815 (44.4)||5296 (46.8)||6056 (51.2)|
|Under-Represented Minority||Not URM||9756 (89.9)||9977 (88.2)||10248 (86.7)|
|URM||1093 (10.1)||1329 (11.8)||1574 (13.3)|
|Submitted K Application||Yes||1862 (17.2)||2386 (21.1)||2561 (21.7)|
|PI on K Award||Yes||1298 (12.0)||1498 (13.2)||1391 (11.8)|
|Submitted R01 Application||Yes||3278 (30.2)||2903 (25.7)||2028 (17.2)|
|PI on R01 Award||Yes||1837 (16.9)||1538 (13.6)||900 (7.6)|
Among all participants, 4,275, or 13%, served as PI on an R01-equivalent award during follow-up. Table 3 shows characteristics according to under-represented minority status. URM matriculants were more likely to be women and to be age 36 years or higher on matriculation; during follow-up they were just as likely to submit K applications and receive K awards, but were less likely to submit an R01 application or receive an R01 award (as PI).
Table 3: T32 Participant Characteristics According to Under-represented Minority (URM) Status
|Total N (%)||29981 (88.2)||3996 (11.8)|
|Age (years) at Matriculation||30 or less||10121 (33.8)||1318 (33.0)|
|31 to 35||13496 (45.0)||1680 (42.0)|
|36 or more||6364 (21.2)||998 (25.0)|
|Gender||Men||15922 (53.1)||1888 (47.2)|
|Women||14059 (46.9)||2108 (52.8)|
|Degree||MD||9844 (32.8)||1439 (36.0)|
|MD-PhD||3224 (10.8)||302 (7.6)|
|PhD||16913 (56.4)||2255 (56.4)|
|Submitted K Application||Yes||5967 (19.9)||842 (21.1)|
|PI on K Award||Yes||3688 (12.3)||499 (12.5)|
|Submitted R01 Application||Yes||7338 (24.5)||871 (21.8)|
|PI on R01 Award||Yes||3882 (12.9)||393 (9.8)|
Time to First Competing R01 Award: Under-represented Minority Status, Age, and Gender
Figure 1 shows that non-URM matriculants were supported as PI’s at a greater rate than URM matriculants.
Figure 1: Time to first R01 award according to under-represented minority (URM) status
Figures 2 and 3 show that for both non-URM and URM matriculants, those who matriculated into post-doctoral programs after age 35 had a lower rate of serving as PI on an R01 award.
Figure 2: Time to first R01 award according to age among those not of under-represented minorities (URM)
Figure 3: Time to first R01 award according to age among under-represented minorities (URM)
Figure 4 shows the combined associations of URM status and gender with rate of receiving an R01 award; female gender and URM status were associated with lower rates of R01 awards.
Figure 4: Time to first R01 award according to under-represented minority (URM) status and gender
Figure 5 shows that among those trainees who received a K-award, non-URM K-awardees were more likely to be a PI on an R01 award sooner.
Figure 5: Time to first R01 award from first K award according to under-represented minority (URM) status
Table 4 shows a Cox proportional hazards model for time to R01 award. The analyses show adjusted associations of age, gender, degree type, matriculation cohort, and K-award (as a time-dependent covariate) with time to award. Men, recipients with MD-PhD degrees, non-URM matriculants, and recipients of K awards were more likely to become PI’s on R01 awards in less time. The model focuses on those 32,739 participants with at least 8 years of follow-up (since few R01 awards were issued earlier, leading to a violation of the proportional hazards assumption). By bootstrap validation the model performed well with an optimism-corrected concordance statistic of 0.786 (where a value of 1 would imply perfect discrimination and a value of 0.5 would imply that the model performs no better than chance).
Table 4: Cox Proportional Hazards Model for First Competing R01 Award among T32 Post-doctoral Matriculants for Those with at Least 8 Years of Follow-up
|Predictor||HR (univariable)||HR (multivariable)|
|Age (years) at Matriculation||30 or less||–||–|
|31 to 35||0.92 (0.85-1.00, p=0.059)||0.82 (0.75-0.89, p<0.001)|
|36 or more||0.46 (0.41-0.52, p<0.001)||0.45 (0.39-0.51, p<0.001)|
|MD-PhD||2.15 (1.91-2.42, p<0.001)||1.67 (1.48-1.88, p<0.001)|
|PhD||1.27 (1.16-1.39, p<0.001)||2.01 (1.84-2.21, p<0.001)|
|2000-2004||1.02 (0.94-1.12, p=0.585)||0.99 (0.91-1.08, p=0.833)|
|2005-2009||0.93 (0.84-1.04, p=0.225)||0.91 (0.81-1.02, p=0.090)|
|Gender||Women||0.71 (0.66-0.77, p<0.001)||0.70 (0.64-0.75, p<0.001)|
|Under-Represented Minority||URM||0.84 (0.74-0.95, p=0.007)||0.83 (0.73-0.94, p=0.004)|
|K Award (Time-Dependent)||Yes||9.52 (8.82-10.27, p<0.001)||10.58 (9.76-11.47, p<0.001)|
In these updated analyses of the outcomes of T32 post-doctoral trainees with MD, MD-PhD, and PhD degrees, we find that non-URM men had the highest rates for R01 application submission and award receipt (as PI). Women, and to a lesser extent, URM matriculants comprised a greater proportion of trainees over time; both women and URM matriculants had lower rates of R01 award receipt (as PI). Serving as PI on a K-awards predicted a substantially higher rate of later support as PI on R01 awards.
I am grateful to my colleagues in the NIH Office of Extramural Research (OER) Division of Biomedical Research Workforce and Division of Statistical Analysis and Reporting for their help collecting these data and conducting these analyses.