The construction of the argument is as follows:
1. 1920s- workers painting watch dials had alarming incidence of a few diseases.
2. cause can be related to radium, leading to discontinuation of use of radium in 1960.
3. as it remains active for 100 years, current watch collectors are still exposed to the radium and should act on it.

Assumption: The radium is not much effective to damage when it is covered with lead boxes or dumped under the Earth. Also no other component of the paint can be found to be as harmful as the radium.

Any statement supporting these two can land you to the answer. Let's check one by one.

A>The surveys also found that the incidence of anemia, loss of teeth, bone fractures and death from cancer among factory workers who painted watch dials with non-luminous paint was not greater than that among the general population.
- This tells that workers with non-luminous paints also faced the incidence which was as likely for them as for the general population. Keep this for now.

B> The surveys also found that factory workers who painted watch dials with luminous paint, the extent of their health problems was directly proportional to the cumulative number of hours for which they had worked with luminous paint till then.
- Exposure to radium is one characteristic and hence this supports the argument to a good extent. Drop this.

C> Barriers of lead or earth provide adequate protection to human beings from harmful radioactive substances.
- This is a straight assumption. Drop this.

D> People who wear watches that have luminous dials and were made before 1960 for more than a decade have an alarming incidence of anemia.
- This strengthens the argument same as B does.

E> Most chemists and laboratory workers who worked with radium for more than a year in the first decade after its discovery in 1898 did not have, for the rest of their lives, a greater incidence of anemia, loss of teeth, bone fractures and death from cancer than the general population.
- This statement states a condition with similar segment working with radium and not having a likelihood greater than that of the normal population for the incidence of the same set of diseases. Keep this and compare with A.

Both A and E weaken the argument. But E compares directly with a similar segment of workers i.e. chemists and lab workers hence this statement weakens the argument better than A by attacking the second assumption. Hence E is the answer.