Comfort is low-variance pleasure. It's pleasure and pain within narrow parameters. When someone is dying, their comfort is our priority, not their capacity for great pleasure; and the opiate medication we palliate them with serves this purpose.
The conservative ethos, broadly speaking, is generally one of preserving a state of comfort at the expense of the possibility for great pleasure or pain (since great pleasure, biologically speaking, tends to have a cost). Risk aversion in humans is a preference for comfort over possibility, and risk aversion is universal, at least among those not currently suffering a great deal.
Assuring the comfort of others is a very humane value.
Here is a problem, though: openness to experience is a highly attractive trait that correlates strongly with certain other highly attractive traits, such as intelligence and youth (and associated neotenic traits). If we are brave and open to experience, we wish to push out of our own comfort zones; think of people you know who choose comfort over adventure, and how you feel about them.
When making moral judgments, we of course wish to display our own highly attractive traits (or even front like we have attractive traits we do not really possess). What better way to signal our own adventurousness than to appear willing to impose it on others?
But is there an important difference between being willing to accept risk ourselves and being willing to impose it on others?
Defining comfort as I have as a state of low-variance pleasure, the ultimate comfort would be the never-born state - pleasure and pain tightly bounded at zero. Creating a person by definition means pulling him screaming out of a state of comfort (negative bliss, as Jim puts it) and pushing him into a state of great risk. Whatever the costs and benefits of this unasked-for adventure, I suspect our feelings about the morality of forcing risk on others function as a (costless, to us) way of signalling our own willingness to take risks.