Based on the results of a few observational studies and meta-analyses [1], many people have drawn attention to the dangers of vitamin A overdose, but let's see how well-founded these claims really are. 

Reading the meta-analysis on vitamin A overdose, it is clear from the data analysed that an unnatural form of retinol, retinol acetate and water-solubilised vitamin A emulsions an intake of as little as 0.2 mg/kg (600 IU/kg) can lead to poisoning in 1-4 weeks, while no documented cases of poisoning have been reported with the nature-identical retinyl palmitate form and up to an amount of 2 mg/kg (6,000 IU/kg, i.e. 300,000 IU/day for a 50 kg person) is not toxic in practice (this daily amount is about 30-60 times the optimal amount, so it can be said that retinyl palmitate is not toxic in practice). This meta-analysis showed that only supplements should only be applied in the retinyl palmitate form and specified the amount of vitamin A causing overdose in the different forms, based on the available data. The meta-analysis also points out that many people supplemented together with vitamin D, and in their case much more vitamin A was needed to produce the negative effects, so the research supports the theory that vitamin D protects against vitamin A overdose. It is clear from this meta-analysis that the natural form of vitamin A (retinyl palmitate) would have to be chronically supplemented with several orders of magnitude more vitamin A than many people already consider problematic, without vitamin D3, to cause a true overdose.   

In another popular 2002 study, [2] vitamin A supplementation increased the risk of bone fractures, but few people mention that it only did so in cases of severe vitamin D deficiency, because it increased pre-existing vitamin D deficiency. If vitamin D levels are adequate, this negative effect is eliminated, and vitamin A even has a positive effect in this respect. Chris Masterjohn wrote a long and detailed article back in 2006 about how vitamin A supplementation only increased bone fracture in this study because it was supplemented without vitamin D, in high doses and in the wrong form (it came from multivitamins, in which the dry form of vitamin A, the unnatural, the easily toxic retinol acetate, is the form typically used, as the safe retinyl palmitate form is oily and thus cannot be properly integrated into powdered capsules/tablets, only retinol acetate is suitable for this purpose). As a source, he provides a wealth of human and animal researches to prove this hypothesis. [3] So, if one supplements vitamin A with vitamin D, or just takes care not to be deficient in vitamin D, there is no negative effect of vitamin A supplementation even at higher doses, and in fact, the two are very effective when supplemented together, as will be shown in reference [4]. 

It is also important to point out that this cross-sectional study published in 2002, although it did find a correlation between vitamin A intake and the risk of pelvic fracture, it was only a correlation, not a causal relationship (since it was an epidemiological study, not an intervention study, which would have been a causal relationship), on the other hand, it is also briefly mentioned in this study that there was no correlation between vitamin A intake and increased risk of fracture in those whose vitamin D3 intake reached an average of 200 IU per day, only in those whose daily D3 intake was below 200 IU (detailed data are not given in the original publication, but are mentioned in the published journal article as being available on request. A publication ten years later also cites these available detailed data, which show that high vitamin A intake was only problematic under 200 IU D3 supplementation.) [5] Almost 20 years have passed since then with several studies completed to determine whether there was indeed a cause and effect relationship, whether high retinol intake/supplementation actually increased the risk of pelvic or any bone fractures. It has been clearly shown to not increase the risk of fractures or pelvic fractures of any kind. For example, here is the result of a study published in 2008 that followed 76,000 women for 7 years: higher retinol intake increased the risk of bone fractures in those with a D3 intake below 440 IU, but high retinol intake reduced the risk in those with an average daily D3 intake of at least 440 IU, even in those with a daily vitamin A intake of more than 25,000 IU, of which at least 5,000 IU was retinol.[6] And in a 2013 Australian intervention study, people who received 25,000 IU of retinyl palmitate per day for 10-16 years had no increase, and even had a slight decrease in their risk of bone fracture compared to those who did not receive supplementation (presumably in Australia, there is enough sunshine and no severe vitamin D deficiency). [7] 

Does vitamin A interfere with vitamin D in any way? 

It is also a common misconception that vitamin A inhibits the absorption/utilisation of vitamin D3, so it is important to take them separately in time. This stems from the misconception that vitamin A pushes vitamin D3 away from its receptors when there is too much of it.   

One can only speculate as to the basis for these claims that vitamin A limits the absorption or utilisation of D3. Perhaps the 2015 in-vitro (human intestinal cell culture) experiment where it was found that high doses of fat-soluble vitamins can somewhat inhibit each other's absorption? However, vitamin D3 and vitamin A did not inhibit each other's absorption at any dose in the study. [8] 

Perhaps it resulted from the fact that there was a 1-day study where subjects were given high doses of retinyl palmitate, calcitriol (1,25-OH-D3) or both in the evening before bedtime and then had their blood levels of calcium, PTH and these two vitamins measured? Even in this case, the two vitamins did not limit each other's absorption, but retinol reduced the excessive blood calcium elevation and PTH depression caused by calcitriol. [9] This in turn shows the protective effect of retinol against the excessive raise in calcium level, i.e. calcification-enhancing, effect of active calcitriol and against calcium outflow from the bones. 

Or perhaps it resulted from the Vitamin D Council (John Cannel) newsletter, expressing concern about D3 supplementation from cod liver oil, which is high in retinol? [10] He also received a rebuttal letter to this letter from several experts, including the head of the test that Cannel had based his concerns on. [11] Masterjohn has also picked apart and pointed out the flaws in his analysis mentioned above.[3]  

In any case, until 2020, there was no human study that could draw a completely clear conclusion about how much vitamin A supplementation is still certainly beneficial alongside D3 and how much is not, or whether taking them together, even at the same time, will worsen or improve their utilization. There were some old studies where they found that if you took retinol, D3 or both at the same time for years at a dose of 100,000 IU, taking them at the same time, it reduced the incidence of colds and caused no side effects, whereas people who took only one or the other developed side effects, so they stopped the trial and reclassified the subjects...[12] And in a randomised placebo-controlled trial (RCT), cod liver oil, which contains many times as much retinol as D3, reduced the incidence of respiratory disease to one-third/one-half. [11] This suggests that there should be no problem with D3+retinol taken at the same time, even if retinol exceeds D3 many times, as it has been shown to be very effective. From then until 2020, there was indeed reason to be cautious about vitamin A (I myself was cautious and suggested supplementation below 5,000 IU, and only if the weekly liver consumption was below 20 dkg, so I practically agreed with the worriers, even if I didn't communicate it as strongly, because of my growing uncertainty.) However, in 2020, a proper intervention study was finally published, which removed all uncertainties: [4] 

The 120 subjects who suffered strokes were divided into 4 groups, with 30-30 subjects in each group. One group received 1 x 50,000 IU retinol per week, another group received 1 x 50,000 IU vitamin D3 per week, the third group received 1 x 50,000 IU D3 + 50,000 IU retinol per week (consumed at the same time!), and the fourth group received nothing (more precisely, placebo). Each group received these once a week for three months. 

Only the group taking vitamin D3 and retinol together had a significant improvement in their condition and inflammation levels, but more surprisingly, their vitamin D status, i.e. their blood 25-OH-D3 (calcidiol) levels, improved (increased) twice as much as the group taking the same dose of vitamin D3 alone. This confirmed beyond any doubt what had already been suspected: The effect of vitamin D3 is enhanced by vitamin A and it is safe to take plenty of both, or even take them at the same time. And what is particularly surprising is that not only did they not limit each other's utilisation, but they improved it, especially vitamin A for D3, meaning that replacing them in a 1:1 ratio is certainly optimal, although as will be discussed at the end, the ratio probably doesn't matter: As long as there is no deficiency of one of them, they only have a synergistic effect, even in huge doses without side effects, while if there is a deficiency of one, taking the other unilaterally increases the deficiency symptoms. 

What is the ideal vitamin A and D3 ratio? Does it even exist? 

It is clear since the 2020 study that the two vitamins are very effective when supplemented in a 1:1 ratio, much better than when taken alone.  This can be considered the optimal ratio. We also know from the studies described earlier that taking 25,000 IU of retinol palmitate per day for over a decade and a half was not without benefit, and we have seen that even taking 440 IU of vitamin D3 per day was beneficial with 25,000 IU of vitamin A per day, but below 440 IU of D3 per day was harmful. This, if you do the math, means that up to a 50:1 ratio (vitamin A to more) is not a problem. 

Data from the earlier famous CARET study, which examined the effects of retinol and beta-carotene supplementation in former asbestos workers (an increased lung cancer risk group), were also evaluated in an analysis published in 2014 and found to be related to vitamin D3 and vitamin A intake, and it was found that vitamin D3 supplementation above 400NE per day reduced lung cancer incidence by half to a quarter if their vitamin A intake was above 5000NE per day, but was ineffective if their daily retinol or equivalent vitamin A intake was below 5000NE. [30] (Presumably, vitamin A intake below 5000NE is suboptimal, so vitamin D3 could not have an effect). If you do the math, this means that you needed at least 12 times as much vitamin A as D3 to get a good effect from D3, while less vitamin A had no good effect. This also pretty much disproves that you shouldn't take too much vitamin A, or that many times more vitamin A than D3 would reduce the effect of D3. 

This confirms what Masterjohn concluded in his 2006 analysis [3]: Whether we are deficient in vitamin D3 or vitamin A, unilateral supplementation of one increases the symptoms of deficiency of the other. However, if we are not deficient in either, then either the ratio is completely irrelevant or the effect is good over a very wide range. As long as we have an optimal supply of them, in any dose, in any proportion, they will eliminate each other's potential side effects while enhancing each other's positive effects. Vitamin D3 (cholecalciferol), once converted into calcidiol and then into calcitriol in the cells or in the kidneys, activates the VDR receptor, while retinol, once converted into retinoic acid in the cell, activates the RXR receptor, and all this forms a complex and triggers the gene expressions to which these two vitamins owe most of their effects. So together they can only trigger it, but they are formed inside the cell, so you can measure their levels in the blood, but you will only see the level of calcitriol produced by the kidney… 

Out of curiosity I looked up breast milk retinol:D3 ratio, based on the reasoning that a ratio that is not harmful for babies, is possible not harmful for adults, either....The retinol content of breast milk per litre ranges from about 1300-2500NE on an average diet (and initially closer to 5000-7000NE), while the D3 content is very low, ranging from 5-80NE/litre. [31] However, if the mother receives 6400NE of D3 per day, the breastfed infant will have the same level as if the mother did not receive D3, but the infant will receive 400NE, which will give a blood level above 100nmol/l, which is good. Sun exposure is about the maximum amount of D3 a mother can get from sun exposure, so it gives a good indication of the ideal level of D3 in breast milk: about 500NE/l. [32] 

In other words, even without vitamin A supplementation and with sufficiently high D3 supplementation, breast milk contains about 2-5 times as much retinol as D3. It is interesting to note that the amount of retinol in breast milk and colostrum is independent of the mother's blood levels of retinol (but depends on the carotenoid and retinol content of her diet). It means that there is an evolutionary mechanism to get the retinol content of breast milk as high as possible, above D3 (since even a sun-loving tropical mother doesn't really get 6400NE, and any food source of D3 has at least 10-40x as much retinol as D3). If you count in the fact that the baby is also sunbathing, the retinol intake is still much higher. 

I think it's pretty clear from what I've explained so far that anyone who gets the 4000 IU or more of D3 per day that either I or Szabolcs suggest shouldn't worry about any amount of vitamin A, but certainly not about 10,000 IU... That said, it's probably unnecessary to go for a 1:1 ratio. However, what we haven't seen is an example of what can happen under a 1:1 ratio. It looks safe above, but what if it's below? Or more precisely, let me ask: How much could be too little vitamin A, which could be aggravated by just a few thousand IU of D3 supplementation? After all, if you don't eat liver regularly and don't utilize carotene-rich foods properly (which is often the case), you'll only get about 1000 IU of vitamin A per day. This is very typical. Is it possible that a 1:1 ratio is the bottom rather than the middle or top of the optimal range? Or do you simply need at least 5000 IU of vitamin A intake per day to be in the optimal range? I raise these questions because several studies have found that children of mothers with higher vitamin D status were more prone to food allergies, [33] while other studies have found the opposite. I think that where higher calcidiol levels were found to be protective against food allergies, maternal vitamin A intake was optimal, whereas where higher serum calcidiol levels were found to increase risk, maternal vitamin A intake was suboptimal. Unfortunately, I cannot verify this, but vitamin A has an important preventive role in the development of food allergies [34], similarly to, for example, how vitamin D3 + retinol together, but not separately, reduces the development of autoimmune diseases. [35] It is not surprising, therefore, that supplementation with vitamin D3 in the presence of suboptimal vitamin A intake predisposes to food allergies, since, as mentioned above, vitamin A and vitamin D increase each other's deficiency symptoms if only one is supplemented but a deficiency of the other is maintained. 

Overall, my advice is to make sure that you get at least as much retinol per week as you get vitamin D3 per week, (including sun exposure) in IU of both. We don't need to do the math exactly, because a multiple of our D3 intake is still perfectly good. If you take at least 4000 IU of D3 every time you haven't had much sun, that's enough to make virtually any supplement or any amount of liver you eat may not cause a problem. It is worth taking into account the high retinol content of liver, which varies between 15-60,000 IU/10 kg depending on the type of liver.  If needed, it is supplemented in the form of retinol palmitate, a natural and safe form of retinol found in the liver. If you don't feel like doing the math, then just supplementing the two vitamins in a 1:1 ratio will only improve the ratio, whether you eat a lot of liver or none at all. Carotenoids can also be a good source of vitamin A for anyone, but their absorption varies greatly from food to food, supplement to supplement and from person to person, so it is difficult to calculate. If you're unsure, don't want to count and worry about it, take a product containing between 4-10,000 IU of D3 every day, and the same amount of retinol palmitate. (Their effects are also enhanced by magnesium and vitamin K1, among other things, and the range of the optimal ratio of these two vitamins, if any, can be pushed even wider). Retinol deficiency is much more common than we think and it is important to do something about it, otherwise D3 supplementation can easily be counterproductive, so don't avoid liver or supplement retinyl palmitate as much as D3.