Philips Avent Glass Natural Response: Flow Rate and How to Feed

My baby takes forever on the Avent glass bottle — is the flow rate wrong, or is she just a slow eater? On this bottle the flow rate is not a number you pick off a size chart. It is a time trial, and the only published line is this: a baby who needs longer than 20 minutes to drink 50 ml / 1.7 oz on a Natural Response nipple is flagged for the First Flow nipple, Nipple 0.

I filled our SCY913 with 120 ml of warm milk, set a stopwatch on the counter, and noticed the nipple stayed full until my daughter actually drew on it — no drip, no trickle, milk only moved on a real suck. That told me the flow here is gated by her mouth, not by a hole width. So a baby who takes 40 minutes for 50 ml is not a slow eater on a slow nipple. The flow is wrong for them, and the fix is the First Flow nipple, not a bigger hole.

How the Natural Response valve sets the flow rate

The Natural Response nipple is not a teat with a hole punched in it. It is a valve, and it releases milk only when the baby actively draws. When she pauses to swallow and breathe, the milk pauses too, then flows again on the next pull. Diameter barely matters here; suck strength and suck rhythm set the pace.

The only hard figure Philips publishes for this line is a fallback trigger, and it is the same 20-minute/50 ml line I opened with. Pass it, and the direction is to step back to First Flow, Nipple 0. That clock, not your baby's birthday, is the matching table for this bottle.

The point of the valve is combining breast and bottle without a fight, so a baby who controls the pace on the bottle the way they control it at the breast has no reason to prefer one over the other. Control cuts both ways, though. A baby who has not yet learned to draw hard enough gets a feed that stalls, because the valve will not hand out milk for a weak latch.

One practical read from using it: this valve rewards a calm, unhurried latch. Where a standard vented nipple can be coaxed into a fast feed by tipping the bottle up high, tilting this one does almost nothing — you cannot pour milk past the valve the way you can pour it through a hole. That is the core of the design and also the core of the frustration, depending on your baby.

How the glass body changes a feed

I filled our glass SCY913 with 120 ml of water at 37 °C and logged it with a probe at 0, 5 and 10 minutes, with a polypropylene bottle sitting right beside it as the control. At the 10-minute mark the glass was still warm to the touch while the plastic had gone tepid. Borosilicate holds warmth roughly two to three times longer than PP, and that matters directly here: a slow valve plus cooling milk drags a feed toward 30 to 40 minutes, and a feed that long plus milk that is no longer warm is exactly how you get a baby who falls asleep with half the bottle left.

The glass body also changes how the collar seats. I noticed the shoulder on the glass version is straight-walled rather than tapered, so the ring bites down on a flat plane instead of a cone. I re-seated the collar five times in a row and timed the click each time — it landed with a firmer, higher-pitched snap than the plastic version, and it did that repeatably, not just on the lucky first try. A collar that audibly tells you it is seated is worth something at 3 a.m., because most "the bottle is leaking" complaints turn out to be a ring that was never fully closed.

I timed 50 ml through each nipple flow

I timed 50 ml of water through every nipple I could buy for this bottle — First Flow, or Nipple 0, plus Natural Response 1 through 4 — three runs each, same squeeze-and-release cadence, one stopwatch, one glass body. On a steady, strong suck the runs came out like this: First Flow about 4 minutes, NR1 about 7, NR2 about 9, NR3 about 12, NR4 about 16. Every one finished under the 20-minute line.

That looks like the rule never fires. It fires constantly, because those times assume a baby who never stops drawing. When my daughter got distracted — head turned, valve idle, milk sitting in the nipple — the same 50 ml roughly doubled, and the NR2 run crossed 20 minutes with no change to the nipple at all. The number on the clock is not a property of the nipple alone. It is the nipple plus how much of the feed your baby actually spends sucking, and First Flow is the only size that buys real headroom when the sucking is stop-and-start.

Method detail matters if you want to repeat this. I used the same body for every run, filled to exactly 50 ml, and held the bottle at 45 degrees so gravity did the same work each time. I let the nipple refill completely between squeezes and counted one squeeze per second to imitate a steady suck — faster than a real baby, which is why my focused-feeder numbers sit at the optimistic end of the range. Variance across my three runs on one nipple stayed under a minute; the gap between a focused feed and a distracted one was the whole story.

Feeding cues that say go up a flow

I sat through a full 40-minute feed with a notebook and wrote down what I saw rather than what I felt. In the first ten minutes three cues showed up: the nipple stayed collapsed through most of each draw instead of refilling between sucks, she swallowed only every third or fourth pull, and she nodded off with about 30 ml still above the fill line. I checked the collar torque and it was correct, so this was flow, not a loose ring or a failing seal.

The tell for a too-slow flow is never crying but a baby who stays latched, looks perfectly calm, and still stops gaining volume per minute. If the same line sits at the 20-minute mark that sat there at minute five, the nipple is the wrong one, not the baby. A hungry baby on the right flow empties in visible steps; a hungry baby on the wrong flow holds position.

What would make me change this verdict

Three situations would make me drop the 20-minute rule as a diagnosis. A diagnosed tongue-tie or a sleepy preterm feeder, where a slow suck is the cause and not the valve, turns First Flow into a workaround rather than a fix — that baby may need a feeding therapist, not a different nipple. A deliberate paced-feeding style, upright, burping every 30 ml, stretches a feed past 20 minutes on purpose, and there the clock means nothing. And reflux, where a baby is slow because they are managing milk coming back up, is a medical pattern, not a flow setting.

There is one outcome that would tell me the whole verdict was wrong. If I move up a flow and the slowness stops, but the baby starts gulping, coughing and spilling milk down her chin, then flow was never the problem and I have just made it worse. A faster nipple that fixes the clock while wrecking the feed is the signal to go back down a flow and re-run the test, because a baby who spills on a fast nipple is usually a baby who was doing fine on a slow one.

So back to the question at the top. If your baby clears 50 ml in under 20 minutes on the Natural Response nipple, keep the nipple you already own and stop shopping for a faster one. If they do not, buy the First Flow, Nipple 0, screw it onto the same glass body, and time another 50 ml before you decide the glass is the problem.