Yes, you could definitely lose weight and still get it. It just improves your odds.
So when docs tell us "Your risk of getting preeclampsia again is X%", the easiest way to think about what that means is probably to imagine that you're playing cards. (Here's another discussion
where cards are the example. I like cards. Sorry about that.)
You've got a deck, with some jokers, and if you draw a joker, you'll get preeclampsia. You won't know if you've drawn a joker until 6-8 weeks postpartum. But you can know roughly how many jokers are in the deck.
In a first pregnancy with no known risk factors for preeclampsia, you've got a deck of 100 cards with 5 jokers in it. That's a 5% chance of preeclampsia.
In a second pregnancy where a doc's given you 20% recurrence odds, your deck of 100 cards has 20 jokers in it.
Losing weight removes some jokers from the deck, depending on your risk. The abstract reports an odds ratio of .7 -- or seven out of ten -- in women who lost weight between pregnancies. So if you've got a 20% chance of getting it again, but then you lose 20 pounds, your risk drops by 6 cards -- .7*20=.14 -- and now you only have 14 cards in 100. (Someone check me!)
I'm trying to make sense of the difference between that -- which seems right to me, but maybe I've got 'thing blindness' and am missing some really obvious statistics -- and their report that in the whole population of women with a history who lost weight intrapartum, this lowered your chances of getting it again from 14.8% to 12.8%. Shouldn't that number be larger with an odds ratio of .7? It might be that the effect is stronger in women who lose more weight, or lose from a higher BMI, but I haven't read the whole full paper yet...
This is a *significantly stronger effect* than the effect of aspirin. LDA has only a .9 odds ratio in the meta-analyses, and 1.0 in the big multicenter trials, which means it would take away 2 cards and leave 18 (at best.)
So losing weight works better than aspirin according to this study. This is not anywhere near as thoroughly evaluated as LDA, yet -- this is a cohort study, not a randomized controlled trial -- but then again, how *would* you randomly assign women to lose weight or gain it or keep it steady between pregnancies?