Are sweeteners and diet drinks safe during pregnancy?

The verdict

Yes, within the acceptable daily intake - and the ADI already includes you. Every regulator that sets one derives it specifically to cover the whole population, pregnancy included. EFSA says so in as many words for aspartame: safe for the general population, including infants, children and pregnant women.

The catch is that the ADI is not the same number in every country. Aspartame is 40mg per kilogram of body weight per day in the EU and Australia, and 50mg/kg in the United States. Saccharin runs from 5mg/kg in Australia to 9mg/kg in the EU to 15mg/kg in the US. Sucralose is 5mg/kg in the US and 15mg/kg in Australia. These are not rounding differences.

None of that changes the practical answer for a normal diet drink habit. You would have to work extremely hard to reach even the strictest of those figures.

Why an ADI is set the way it is

An ADI starts from the no-observed-adverse-effect level in animal studies - the highest dose at which nothing happens - and divides it, conventionally by 100. A factor of 10 covers the possibility that humans are more sensitive than the test species, and another factor of 10 covers variation between humans, which is the slot that pregnancy, infancy and illness sit in. So the number is already two orders of magnitude below the level where anything was observed in a rat.

What happens in the body differs by sweetener, and this is worth knowing because it drives the one genuine exception. Aspartame is a dipeptide: it is digested into aspartic acid, phenylalanine and a small amount of methanol, all of which occur in ordinary foods in larger quantities. Phenylalanine is why aspartame carries a label warning, and why aspartame is genuinely unsafe for one group - people with phenylketonuria, who cannot metabolise it. If you have PKU, or maternal PKU is being managed in your pregnancy, that is a conversation with your metabolic team, not a website.

Sucralose, acesulfame potassium and saccharin are largely not metabolised at all. They are absorbed to varying degrees and excreted essentially unchanged. Steviol glycosides are hydrolysed by gut bacteria to steviol, absorbed, conjugated in the liver and excreted.

All of them cross the placenta to some extent - acesulfame, cyclamate, saccharin and sucralose have all been detected in fetal circulation and amniotic fluid in human samples. That fact is often presented as alarming on its own. It is not, in itself, a finding of harm; caffeine, paracetamol, glucose and every amino acid you eat cross the placenta too. It is the dose that decides.

How much is too much

The FDA publishes the most usable version of this: how many tabletop sweetener packets a 60kg adult would have to consume in a day to reach the ADI.

• Aspartame: 75 packets
• Acesulfame potassium: 23 packets
• Sucralose: 23 packets
• Saccharin: 45 packets
• Rebaudioside A (a steviol glycoside): 27 packets
• Advantame: 4,920 packets

In drinks rather than packets, a 330ml can of diet cola contains roughly 180mg of aspartame. Against the strictest ADI in the world - 40mg/kg, so 2,400mg for a 60kg adult - that is about thirteen cans a day before you reach the limit. Against the US figure it is sixteen.

There is a second number on a diet drink that is much easier to hit, and it is not the sweetener. The NHS puts a can of cola at 40mg of caffeine, against a 200mg daily caffeine limit in pregnancy. Diet and regular versions are not usually different in that respect. Five cans is your whole day's caffeine, with no coffee or tea in it. In practice the caffeine budget will constrain your diet cola long before the sweetener does.

Natural sweeteners are not a separate category

Stevia is regulated exactly like the synthetics. Its ADI, 4mg/kg expressed as steviol equivalents, was set by the Joint FAO/WHO Expert Committee on Food Additives and adopted by the FDA, EFSA and FSANZ alike - and it is the lowest of the widely used sweeteners on a body-weight basis. The FDA also draws a line most shoppers do not know exists: highly purified steviol glycosides are permitted, but whole-leaf stevia and crude stevia extracts are subject to an import alert and are not permitted as sweeteners at all. Monk fruit extract has no specified ADI in the US or Australia.

If you have been drinking a lot of diet drinks

There is nothing to undo and nothing to test. No authority treats sweetener intake as an exposure event with a follow-up.

There is one piece of evidence worth knowing about honestly, because it is the reason the WHO looked at this at all. Single prospective observational studies have reported associations between non-sugar sweetener use in pregnancy and outcomes in the child, including higher infant body mass index, asthma and allergy, and preterm birth. These are observational, they are inconsistent, they cannot separate the sweetener from the dietary pattern that goes with it, and no regulator has changed an ADI because of them. They are the basis for a research gap, not for a warning.

The sensible move is not to switch back to sugar. It is to notice how much of your fluid intake is fizzy and sweet at all, and to displace some of it with water - a recommendation that would be the same if the drink were sweetened with cane sugar.

Where the guidance genuinely differs

This is one of the widest regulatory spreads on the site, and it exists because agencies re-evaluate at different times and read the same animal data differently.

The ADIs do not match. Aspartame: 50mg/kg (FDA), 40mg/kg (EFSA and FSANZ). Saccharin: 15mg/kg (FDA), 9mg/kg (EFSA, raised from 5mg/kg in November 2024 after a full re-evaluation), 5mg/kg (FSANZ). Sucralose: 5mg/kg (FDA), 15mg/kg (FSANZ). Same molecules, same literature, three different numbers, none of them recently reconciled.

Cyclamate is the real country-level split. It is prohibited in the United States. It is a permitted food additive in Australia and New Zealand with an ADI of 11mg/kg. In Canada it may not be used in food at all, but may be sold as a non-food tabletop sweetener - and when it is, the Food and Drug Regulations require the label to state that the sweetener should be used only on the advice of a physician. If you are in Canada and you have a jar of cyclamate sweetener, that instruction is aimed at you.

The WHO takes a different line from all of them. Its 2023 guideline advises that non-sugar sweeteners not be used as a means of achieving weight control or reducing the risk of noncommunicable disease - a conditional recommendation, based on a review finding no long-term benefit for body fat and possible undesirable effects with long-term use. That is not a safety ruling and it does not contradict any ADI. It is a statement that they do not do what people buy them for.

The UK sits with the EU on the science and with nobody on the messaging. The NHS states that all approved sweeteners are a safe and acceptable alternative to sugar, lists eight permitted in Great Britain - acesulfame K, aspartame, erythritol, saccharin, sorbitol, steviol glycosides, sucralose and xylitol - and does not mention pregnancy at all on the subject.

If you want a single rule that satisfies every jurisdiction: stay inside the lowest published ADI, avoid cyclamate tabletop sweeteners, and count the caffeine in the can rather than the sweetener in it.