From my experience, one or two instances aren't going to be damaging if you're generally on track most of the time and people were pretty hyped up as you took the break. How you return is just as important as how you take that break. If you say you'll be back in a month and you fulfill that promise, your next break isn't going to be a problem.
But there's a lot of things that drive readers to a page. If the majority are relying on habitual surfing (to your site) you're probably in a more vulnerable position than those who have a lot of their readership on other communication channels whereby they will be alerted to your new posts. (One problem I saw was with some comics that generally ended or went on hiatus that the readers I shared with those sites did not always come back to my page. It was vexing but just the way it rolled at the time.)
The question is whether this is a brief break or one wher eyou don't know the end? If it's indefinite, try for an opt-in mailing list . You may also want to query your readership now on how they get alerted to what's going on.
Also -- do you have the option simply to drop your posting frequency per week? That might allow you more flexibility and less concern about missing an update. You'll not miss updates, you'll just have more time in between them .
Retention: Oh yeah, you lose people unless you launch right away.
I was scolded by a few readers in my final installments of the first serial for not being ready to go immediately with its sequel but not much I could do. I had a lot of life things that took priority. (Sorry but job >> free serial.)
Statistics-wise for me, the denominator for the second serial I started was only going to be on a scale with those who finished the first. I haven't looked at the website (and the impact of three months off between serials).
Wattpad definitely saw a drop.
I have no hard numbers on Wattpad but guesstimates. I think I saw anywhere from 20-40% of people who finished the first book move on to the second. I'm not sure as they don't offer linked data. I can only presume that people would not start reading a "volume 2" without going to volume 1.
I haven't run numbers on the website yet to know if the drop is similar in scale. Also, not sure how to factor for differences in blocking nuisance traffic. Wordpress' Jetpack picks up a lot of crap so I'll have to dig into analytics to get a better sense of what's going on.
The website attrition is partly my fault, however, as I had no mailing list to reengage my audience when I was back. The majority of readers do not use Facebook, Twitter, Tumblr, Deviantart, etc. They were direct visitors for the most part.