Table 4.

The BMD recommendations modeling hepatic endpoints for Osweiler et al. (1993) and Jennings et al. (2020).

						BMDS Recommendations and Notes
	Study	FUM dataset	Endpoint	Response means	Response variability	Exponential 3	Exponential 5	Hill	Linear	Power
	Osweiler et al. (1993)	5, 31, and 148	AST	70.5, 79.5, 392.3*	2.1, 10.1, 136.7*	Unusable	Questionable	Questionable	Questionable	Questionable
						Did not successfully execute.	Zero degrees of freedom; saturated model	Zero degrees of freedom; saturated model	Nonconstant variance test failed (Test 3p-value < 0.05)	Zero degrees of freedom; saturated model
							Nonconstant variance test failed (Test 3p-value < 0.05)	Nonconstant variance test failed (Test 3p-value < 0.05)	Goodness of fit p-value < 0.1	Nonconstant variance test failed (Test 3p-value < 0.05)
							Control stdev. fit > 1.5	Control stdev. fit > 1.5	Control stdev. fit > 1.5	Control stdev. fit > 1.5
	Osweiler et al. (1993)	5, 31, and 148	GGT	17.5, 20.5, 84.3*	1.71, 3.67, 14.45*	Recommended: lowest AIC	Questionable	Questionable	Questionable	Questionable
						BMD − 18.2	Zero degrees of freedom; saturated model	Zero degrees of freedom; saturated model	Zero degrees of freedom; saturated model	Goodness of fit p-value < 0.1
						BMDL − 8.2
						BMDU − 41.3
	Osweiler et al. (1993)	5, 31, and 148	LDH	952.8, 970.7, 1811.1*	55.9, 51.3, 255.8*	Recommended: lowest AIC	Questionable	Questionable	Questionable	Questionable
						BMD − 64.3	Zero degrees of freedom; saturated model	Zero degrees of freedom; saturated model	Zero degrees of freedom; saturated model	Goodness of fit p-value < 0.2
						BMDL − 34.7
						BMDU − 134.5
	Jennings et al. (2020)	5.0, 26.7, 40.7,	SA:SO	0.07, 0.08, 0.2,	0.01, 0.05, 0.12,	Questionable	Questionable	Unusable	Unusable	Unusable
		67.7, and 108.8		0.3, and 0.8*	0.08, and 0.45*	Control stdev. fit > 1.5	Lowest dose/BMDL ratio > 3	Control stdev. fit > 1.5	Lowest dose/BMD ratio > 3	Control stdev. fit > 1.5
						Constant variance test failed (Test 2p-value < 0.05)	Lowest dose/BMDL ratio > 10	Constant variance test failed (Test 2p-value < 0.05)	Lowest dose/BMD ratio > 10	Constant variance test failed (Test 2p-value < 0.05)
							Control stdev. fit > 1.5	BMDL does not exist	Goodness of fit p-value < 0.1	BMDL does not exist
							Constant variance test failed (Test 2p-value < 0.05)		Control stdev. fit > 1.5
							BMD/BMDL ratio > 3		Constant variance test failed (Test 2p-value < 0.05)
							BMD/BMDL ratio > 20		BMDL does not exist

*Reported to differ from corresponding control.