When writing either a review section of a paper or an entire review paper, it is extremely important to be clear about which contributions came from which paper, and which form of evidence they are: does that paper have direct evidence on that point, rely on other papers/rely on an assumption, or develop the point as speculation/theory/explanation?
Obviously, you should cite any of these that you use, but you should cite them differently. For instance, if you have an article by author X, X might condense the results of previous literature (possibly citing authors Y and Z), present their own experiment, and discuss the theory and possible implications of their own experiment and interpret it in light of the literature. You would want to distinguish the three.
X emphasizes that the previous literature finds a strong relationship between concept 1 and concept 2, and considers the work of Y and Z to be "extremely strong evidence that causality runs both ways" (X, p. 876).
X finds [shows, provides evidence] that [experiment results] [in context].
X argues [concludes, suggests, wonders, hypothesizes...] that [explanation put forth].
I'm guessing you were already thinking about a certain hypothesis, and then you saw X's paper and realized that they had thrown in that hypothesis at the end. Definitely cite them. Even if you had been looking at X's paper for the experimental results, you should also re-read what their discussion/conclusion says, and if those are relevant to the discussion, you should acknowledge the ideas in some way. (When would their discussion or conclusions not be relevant? Perhaps if you're interested in something that they had to account for on the way to their focus.) It could even be a simple mention just to acknowledge that they said it, too.
Many scholars consider concept 1 to be a likely explanation (cf. X, Y and Z, Q).
It's OK if you conclude your review with an interpretation that's similar to what X suggested, as long as you have reviewed articles suggesting other possibilities and you reference as you're concluding that it's similar to X.
Overall, there is mounting evidence that, as X suggested, [explanation].
Your contribution is a consideration of additional material, synthesis of ideas, and your own opinion/emphasis on what's important. For the purposes of publishing, if X has already covered similar area, your review paper may not be extremely interesting or important unless you can show why the overall question/field matters or that people are still working under alternative assumptions or not yet convinced.