Disproportionately in this case means a higher percentage of women 'failed' the test than men.
So say Koch hires 100 people - 80 men and 20 women. 10 of the men fail the test and 10 of the women fail the test. Then men had a 12% failure rate while the women had a 50% failure rate - women are disproportionately failing. That by itself is not a problem. The problem comes when the test does not accurately assess a candidates ability to do the job. The first problem is isokinetic testing is not designed for competency testing as it does not simulate real world abilities. It is best used for diagnostics and rehab - ie identifying the problem muscle and quantifying development. The second problem is "were the 'standards' set for truck drivers, or were they based off of a entirely different profession?".
It's like making potential hires take a mental math test. Do drivers need to be able to do math? Yes. Do they need to be able to do it in their head? Sort of, but not really. Do they need to be able to multiply 3 digit numbers quickly? Nope. So why would you give potential drivers a test that was designed for stock traders? The test itself does not indicate an ability to do the job. If the results of such a test disproportionately eliminated a group the company is practicing discrimination.