Just like my HYPERTROPHY ARTICLE, this is an opinion piece as opposed to an attempt to systematically answer a question. In this instance, the question is “should weight loss diets be lower fat or lower carb?”.
The answer, strictly speaking, is that it doesn’t matter. Provided protein and energy intake is matched, and presuming perfect adherence, there is no meaningful difference in results between diets of either persuasion.
Despite that, I mentioned in THIS article that I have a preference for lower fat diets as a starting point. I have a few reasons for this belief and in general my opinion is that given the lack of a compelling advantage to lower carb diets, I think there are some points in favour of lower fat diets (/those that preserve carbohydrate intake) which make them appealing.
That said, given that both approaches are equally effective when done right, and given the variability in individual food preferences and satiety responses to diets, low carb diets are still a perfectly viable option. Were I to be taking a shot in the dark, my first shot would be towards reducing dietary fat (specifically added fats) and promoting intake of foods which were carbohydrate sources (vegetables, fruit, wholegrains), but I’m certainly not opposed to lower carb approaches especially if there were indications on the part of the individual that that is what they may prefer or thrive on. Dietary approaches should be personalised to aid adherence and wellbeing.
Now first, to reiterate, there is no systematic advantage to lower carbohydrate diets for weight loss and modelling has shown a meaninglessly small advantage in the other direction (ie to carbs). So presuming people will eat what you give them, the choice is not an overarchingly important one.
Except that there are a number of factors that influence adherence to and success of diets which are related to food choices and therefore the nutrient composition of the diet.
The first concept to consider is the energy density of the diet. Energy density means (as you’d expect) the amount of energy per gram/mL of food/fluid consumed. The macronutrients contain different amounts of energy per gram – protein and carbs contain about 4kcal/g (there’s some conjecture about protein, but it’s near enough) while fats contain about 9kcal/g.
As I stated at first, I’m presuming protein intake is unchanged, whether you elect to eat a higher or lower carb diet. For a given amount of energy, the weight of food consumed will be greater if that energy comes from carbohydrate. (Strictly, energy density has been examined independently of macronutrients and been observed to have an effect in and of itself, so whilst manipulating macronutrient composition manipulates energy density, they are not synonymous).
The other big determinant of energy density is the water content of the food consumed (as water contains no energy, but obviously weighs 1g/mL). Solid foods with high water content, such as fruits and vegetables, also are typically higher carbohydrate and lower fat. I’m not entirely sure whether that is coincidence or just biological reality because fats are hydrophobic, but in practical terms the important thing to consider is that the foods that yield the fewest calories per bite are carbohydrate-containing, and the best way to reduce the calorie density of a diet is to reduce its fat content.
The next question is, of course, why reduce the energy density of a diet? We obviously have biological systems that regulate hunger. They are enormously complicated and are partially regulated by acute and chronic levels of energy intake, but that’s one part of a much more complex equation. The weight and volume of food consumed also has effects on appetite, with distension of the stomach one indicator of “fullness”.
There is plenty of literature to support reductions in ad libitum energy intake with lower energy density diets. There’s also some evidence for consuming strategically low energy density snacks, or low energy density preloads (such as broths) before meals to enhance fullness/satiety.
In a study in which energy density was covertly manipulated over 14 days participants DID modulate their intake of food weight to compensate for the reduced ED of the diet, but not enough to completely mitigate the decrease in energy intake from consuming a low ED diet. Whether long-term reductions in intake can be sustained PURELY through a reduction in energy density, I’m uncertain, as other drivers of hunger might cause a further compensatory increase in the volume of food consumed or reduction in energy expenditure and so on, but it does support the notion that lower energy densities would support lower energy intakes (and vice versa).
Similarly, in a study using high/low energy density preloads before a buffet-style lunch, those who consumed low energy density preloads reported more less fullness/more hunger following the preload and consumed slightly more at the ensuing meal, but not enough to fully offset the reduction in intake from the low ED preload. Again, given that there was compensatory behaviour at the subsequent meal, and by the end of the day energy intake was not significantly different between the groups, I don’t consider this strong support for making ONE low energy density substitution in the diet.
However, given that the high ED condition ate more across the two meals (preload and lunch) I would say that it supports the notion that high energy snacks do not necessarily attenuate hunger enough to be “worth it”.
Further, given the general support for low energy density diets overall in reducing intake one could say that overall dietary context is more important and that where diets are structured such that compensatory behaviour can occur, single behaviours as opposed to overall changes in diet patterns are relatively ineffective. But that also seems very obvious.
Bringing all of that back to the question at hand – the reason I think this is important is that while many people are willing to engage in conscious “dieting” behaviours strategically and on a short term basis (think things such as following a meal plan, counting calories etc), most people do not want to continue those behaviours for life.
However, for weight loss to be sustained there needs to be a commensurate reduction in long-term energy intake to the appropriate level for the new bodyweight/activity level (which is part of why it is so difficult to keep weight off).
I’d go as far as to say that weight loss itself is actually not very difficult, in that relatively short-term efforts can yield quite impressive results. But sustained weight loss IS, and planning maintenance behaviours is one of the most important aspects of it. Choosing a dietary structure that facilitates lower energy intakes and feelings of fullness during ad libitum eating is doubtless helpful in this respect, and from a practical perspective it also gives you more leeway for making portion errors when the extra calories yielded by an accidentally heaped spoonful are lower.
So, for sustained weight loss I think helping people move towards a low energy density diet pattern is probably helpful because it will likely help facilitate long-term reductions in energy intake without as much conscious input. This is related to the concept of focusing on behavioural factors I spoke about HERE.
It is also true that low energy density diets also tend to be higher “quality” – remember what I said about fruit and vegetables being low energy density? And so you might argue that dietary “quality” (meaning minimal processed/hyper-palatable foods high in fat, salt and sugars in combination) plays a role also, and that’s likely true. But you would have to contrive a pretty ridiculous diet for it to be low quality and low energy density simultaneously to prove that point semantically.
That is supported by this analysis of a US population, where low energy density diets were comprised of more fruits and vegetables and less added fat. Added fat was inversely related to the amount of food/drinks from the major food groups and positively related to consumption of low nutrient density (aka “junk”) foods. Higher energy diets contained less fibre, protein and micronutrients.
For the sake of completeness, none of the above is to say that by virtue of being typically more energy dense, low carb diets are rubbish for controlling hunger.
Firstly, anecdotes abound of much “steadier”/reduced hunger with low carb diets (although for what it’s worth I could find just as many saying the opposite), and probably more importantly, experimental evidence suggests that they can foster spontaneous reductions in energy intake too.
Ketogenic diets do appear to have some advantages for controlling hunger and spontaneously reducing energy intake (keto diets are at the extreme end of the low carb spectrum). See here and here.
By virtue of limiting carbohydrate intake, any low carb diet (from “moderate” through to keto) also reduces intake of hyperpalatable foods that are easily overconsumed (such as cookies, ice cream, pizza, cake etc), so neither is that advantage exclusive to lower fat diets.
So, given that many individuals may prefer this approach, they’re absolutely fine in that respect. But my first point in favour of lower fat diets as a starting point is that for most people that dietary pattern will aid in sustained reduction of energy intake.
The second, related, concept to consider is associations between foods and nutrients and weight loss/control. Again, this discussion needs to be couched in a huge degree of scepticism because there are so many factors that could contribute to the relationships we observe other than fat/carbohydrate content and energy density, such as protein content, overall dietary quality and replacement effects (consider that those who eat the most fruit are likely replacing less favourable snack foods etc), socioeconomic status and so on.
However, there are fairly consistent associations between fruit/vegetable intake and wholegrain intake and bodyweight control, all of which are foods that yield mostly carbohydrate.
Given that (as I said in my last article) we eat foods as opposed to nutrients (ie, we don’t eat “a carb” for lunch), I think being at least mindful of which foods seem to comprise “successful” diets is valuable, provided we are conscious of the numerous potential explanatory factors. Even more true is that we don’t just eat “foods”, really, we follow an overall dietary pattern, so considering the intake of foods in context is important too.
Likewise, as I also said in my last article, the relative absence of highly processed foods in the diets of those with the best bodyweight control is likely just as important.
Again, for completeness, there are also associations between consumption of both dairy and nuts and lower bodyweight, neither of which are strictly low fat (nuts especially). As with energy density, at least part of the pattern is likely attributable to diet quality.
Plus, changes such as reducing fat/increasing fruit intake are indicative of attempts at improving dietary health, so the associations between either behaviour and bodyweight (as observed in this study) could just as easily reflect “effort” as opposed to a food-specific effect.
This paper makes some interesting points, among which being that reduction in fat intake appears to lead to modest weight loss, however with greater success where dietary advice is focused on food composition (as opposed to simply saying “reduce fat intake”) and also with greater success when combined with increased fiber intake. To state the obvious, promoting reduced dietary fat through food choices whilst increasing fiber intake is likely to mean advocating swapping higher fat foods for fruit, vegetables or wholegrains.
In practice, how this matters to me is that in looking to replace lower quality, easily overconsumed foods with better options (those being largely vegetables, fruit and wholegrains) I think we have a reasonable basis to say that the best food choices are often lower fat.
Or, maybe more fairly, diet patterns that contain large amounts of fruit, vegetables, wholegrains and dairy, and are lower in junk/processed foods seem to help with bodyweight control. Carbohydrates will by definition form a reasonable component of diets that fit those stipulations, and fats (especially added fats) are likely to be reduced.
The next reason I prefer lower fat diets as a starting point is for performance. It is pretty uncontroversial to say that higher carbohydrate diets are preferable for athletic performance in nearly every field. Attempts at cyclical nutrition aiming to maximise adaptations to low carbohydrate intake such as higher fat-burning capacity and sparing of carbohydrate stores during performance have been pretty underwhelming, as the adaptations are not preserved long-term, carbohydrate metabolism is impaired with lower carb diet phases, and sustaining sufficiently productive training with low carbohydrate intake is hard.
I have only trained a few TRULY elite athletes, and only one athlete who was legitimately interested in endurance exercise performance (where carbohydrate requirements are highest), but even for recreational trainees there are benefits to maintaining at least adequate carbohydrate intake. Admittedly, for recreational weight trainees or recreational sports people, “adequate” carb intake is quite low (I often start with 2-4g/kg, depending on the amount of activity people are doing).
However, in a weight loss context, especially where energy expenditure ISN’T super high by virtue of doing a lot of exercise, that level of carbohydrate consumption can still represent a significant portion of the energy allotment.
So, again, a lower carb diet is a reasonable option in a weight loss context, but in order to sustain performance (which I am interested in, and I believe most people pursuing weight loss in a recreational context should be too), there is a minimum carbohydrate intake that I wouldn’t advise falling below.
The MOST important factor for weight loss is adherence. Both on a short-term scale, and for the long term sustainability of weight loss as I have already alluded to. Intermittent lapses in adherence describe weight loss plateaus, rather than metabolic adaptation, and long-term erosion of adherence to caloric restriction likely describes regain.
A recent trial compared energy/protein-matched low fat and low carb diets for weight loss finding (unsurprisingly) that neither was more effective than the other – see this writeup covering it. Participants in the low carb group began with <20g/d of carbs, before being allowed to titrate up intake to the lowest level that they found sustainable. For a large majority that was considerably higher, with an average carbohydrate intake in the low carb group of 132g/d (which is admittedly still low). The low fat group settled at an average intake of 57g/d, which is quite low, but not at the point that food tastes like cardboard.
When considering what is a “sustainable” dietary pattern, considering individual preferences and the preservation of food choices and social eating opportunities are paramount. Having carbohydrates arbitrarily restricted beyond a point appears to be hard for many individuals, whilst a lower fat diet can mostly be sustained mostly through reducing added fats and smart food selection.
However, as I’ve already said, plenty of individuals did great on low carb diets (and plenty didn’t do great on low fat diets – see the graph of individual responses in the article I linked). As the author of the paper states in the Q and A, once the foundations of a good diet are in place, personalising to maximise adherence is the way to go.
So, having written all of that perhaps the fairest summary of my position isn’t ACTUALLY that low-fat diets are the way to go for weight loss, in that it is not the restriction of fat itself that I think is most important.
Rather, I think that when we consider the components of diets that successful promote reduced ad libitum energy intake (basically an interrelated jumble of low energy density, high fiber, reduction in processed foods and a basis in fresh fruit and vegetables, wholegrains and dairy, with adequate protein as a must) the resultant “pattern” is one of low to moderate fat. Improvements in dietary quality often entail a reduction in dietary fat, and giving some consideration to the ease with which people can restrict a macronutrient is important too.
However, as MUST be said, once the diet plan is set, there isn’t a difference in results from either a high or low fat/carb diet. The differences in results are really down to adherence, and so whilst I believe a rational starting point is improving dietary quality through the focuses above, beyond that point I think low carb diets can be an excellent option and if they are preferred by the individual then by all means have at it.
A few additional notes
- Perceived simplicity of a dietary strategy is related to adherence. This is (I suspect) one reason why “carbs are bad” or “x food is bad” or “just do paleo” approaches actually work very well. If somebody just finds it easier to think in terms of excluding certain carbs (or fats, or whatever) then that is also not a terrible basis upon which to start a dietary intervention.
- Given that the carb:fat ratio doesn’t actually matter, provided the energy reduction is achieved one way or another, there is of course room for moderate intakes of each. However, a diet that deliberately lowers “added” fats and focuses on lean meats/fish and dairy, with a large basis of vegetable, fruit and wholegrain intake will still contain some fat and it’s not hard to eat a diet that is 25-30% fat without every adding fat sources to meals or choosing to snack on nuts.
- Every attempt at dieting to lose weight, even if unsuccessful, is valuable if only as an information-gathering exercise. If you, or a client, attempt to lose weight on a low fat (or low carb) diet and dislike it or find it unsustainable, that information is useful for later strategic planning. Do not consider an immediate failure to achieve a result as an indication that you cannot succeed, but it may be useful means by which to exclude avenues for future attempts.
Now, if you prefer low carb diets, here is a three point summary of how to minimise the drawbacks of low carb dieting and take advantage of the associated with higher carbohydrate foods/diets above.
1 – Consume a large amount of non-starchy vegetables. They contain relatively negligible digestible carbs and energy, lower the energy density of the diet and provide large amounts of fibre.
2 – Consume adequate protein and adequate carbohydrate to support performance. It may be worth considering the timing of carbohydrate intake to facilitate training performance or following a cyclical approach, but at the least consume a broadly appropriate amount.
3 – Minimise addition of fats to meals. Consuming meats/fish/poultry, full fat dairy and nuts along with a large amount of non-starchy vegetables is a great basis for a low carb diet. Adding butter/oil to everything isn’t. I tend to think most of the fat in diets (whether lower or higher fat overall) should be “incidental” – meaning it is intrinsic to the staple foods of the diet as opposed to added in the process of cooking/eating.
Oh, and no matter what you do I think you should eat a decent amount of protein and lift weights.
If you want to hear me prattle on about training, you can listen my podcast – it’s here. Leave a review on iTunes and I’ll read it out on air.