Weighing food: how to use a kitchen scale

Because the insulin dose is based on grams of carbohydrate, and those grams come from weighing food as precisely as possible. The eye judges volume well but weight poorly, especially with the denser foods. The scale turns "a portion of rice" into a number you can actually use in a calculation [1].

Without a figure from the scale, the same plate can mean 30 g or sometimes even 60 g of carbohydrate. The difference is not theoretical: at the same insulin dose, a difference of 20 g of carbohydrate leads either to hypoglycemia or to hyperglycemia after the meal [2].

The scale helps you find the weight very precisely, not the carbohydrate. The carbohydrate value comes from the composition table or from the label and is applied to the weight you measured [3].

The second gain from using a scale is that it trains your eye. After a few weeks of weighing you will notice that your visual estimates come closer and closer to reality [1]. The scale then stays useful mainly for new foods and for very concentrated ones.

Choose a digital scale that displays single grams, not steps of 5 g. Then repeat the same four steps at every weighing:

• place the scale on a hard, flat surface, never on a towel or on the edge of the sink;
• put the empty bowl on top and press the "TARE" button, to bring the display back to zero;
• add the food and read the figure, which is now the net weight, without the weight of the bowl;
• press "TARE" again before every new ingredient, so that you weigh the rice and the sauce separately.

Bringing the scale back to zero after a container has been placed on it is called taring. The button marked TARE applies the tare function.

Write the figures down straight away, before you add the next ingredient and above all before you start cooking. A figure written at the moment of weighing no longer depends on your memory [4]. Check the scale from time to time against a reference item of 500 g or 1 kg. On some models, a weak battery can be responsible for readings that drift further and further from reality.

The golden rule is to weigh the food in the same state as the one given in the table [5]. If you use the row marked "raw", you weigh it raw, and if you use the row marked "boiled", you weigh it boiled. Both options are correct, provided you do not mix them. The usual mistake is to weigh cooked pasta and apply the figure for raw pasta [3].

In practice, weighing raw is more accurate for pasta and rice, because the weight after boiling depends on the cooking time and on the water absorbed. Weighing cooked is more convenient when you eat from a shared pot or in a restaurant. Choose one habit and keep it, so that you can compare your meals properly. In most situations you will probably weigh the food ready to eat.

They absorb water and hold on to it, which dilutes the carbohydrate. The dry starch in pasta and rice swells on contact with hot water and soaks it up. So 100 g of raw pasta or rice reaches 250-300 g after boiling. The amount of carbohydrate stays roughly unchanged, only water has been added.

That is why the figures look so different for these two food categories: 70-75 g of carbohydrate per 100 g raw, against 25-30 g per 100 g boiled. Preparation therefore remains one of the important sources of error in estimating carbohydrate [3].

Boiled potato also changes weight a little, and meat loses between a quarter and a third of its weight during cooking [6]. That loss does not matter for counting, because there is no carbohydrate in meat. It is simply an example of weight changing during cooking. Frying drives water out of the product and therefore concentrates the carbohydrate.

You weigh only the edible portion, that is, what actually ends up on the plate. Tables give carbohydrate values per 100 g of edible portion, without peel, stones, core or bones [7]. The difference is far from small: the peel of a banana is about a third of the fruit, an orange peel about a quarter, and an apple core roughly a tenth. For the same reason, portion lists work with net weights, not with the whole fruit [8].

You have two ways of dealing with foods that come with peel, stones or bones. Either you peel first and weigh afterwards, or you weigh the whole food, then the leftovers, and take the difference. The second method also works for grapes on the stalk or for a melon. With meat on the bone do not overcomplicate things, because meat brings no carbohydrate. The point is to stay attentive to the foods that do carry carbohydrate.

The table tells you how many grams of carbohydrate a food has per 100 g of product. You multiply the weight you measured by that value and divide by one hundred. For example, 200 g of well-cooked rice (pilaf), at 25 g of carbohydrate per 100 g, means roughly 50 g of carbohydrate. The same arithmetic also sits behind carbohydrate portions.

The hard part is choosing the right row: raw, boiled, baked or fried, with or without peel. Large databases such as FoodData Central and the European national tables have a separate row for each [7]. Be careful with American sources: there, carbohydrate includes fiber, whereas on the EU label the "Carbohydrate" line does not [9] [10].

Neither the table nor the label is perfect, and their errors add to those of your own estimate [11].

As long as possible at the start, then only periodically. In the first few weeks it is worth weighing everything that contains carbohydrate, especially the meals you repeat often. That is how you build a set of mental images of what 60 g of bread or 200 g of potatoes look like on your plate. The effort makes sense, because about half of adults with type 1 diabetes say that carbohydrate counting remains a difficulty [12].

After a while, visual estimates become less and less accurate [13]. That is why it is very important to go back to the scale for a few days every few months, as a kind of recalibration of the eye [14]. Always weigh new foods and new recipes, especially the ones very concentrated in carbohydrate. Foods where a few grams more can meaningfully change the insulin dose are bread, sweets, breakfast cereals and dried fruit.

Yes, it can happen, and the risk deserves to be taken seriously. Disordered eating is more common in patients with type 1 diabetes [15]. Having to weigh food every day adds to the other pressures the disease brings, and the carbohydrate counting method itself has been associated with the risk of disordered eating [16].

A rigid meal plan and strict tracking of every gram of food can feed anxiety in people who are already vulnerable. The scale does not create the problem, but it can fuel it. The warning signs are the following:

• guilt after meals;
• refusing to eat anything that has not been weighed;
• avoiding meals out;
• thinking about food constantly;
• skipping insulin doses, the most serious sign of all [17].

If you recognize them, tell your team. The scale is a measuring instrument, not a judge.

The foods that bring a lot of carbohydrate in few grams of product always go on the scale: bread, pasta, rice, potatoes, breakfast cereals, sweets and fruit. For bread, one extra 30 g slice means roughly 15 g of carbohydrate, which is a real change of insulin dose [2]. It is precisely with these foods that visual estimates go wrong most often, in the direction of underestimation [1].

You can estimate non-starchy vegetables by eye. Here a mistake in the estimate does not matter much. For example, 50 g of extra salad means around two grams of carbohydrate. Estimating with the hand works reasonably well for vegetables, but worst of all for grain foods [18]. Still weigh starchy vegetables, pulses and nuts, which are usually eaten in larger amounts than expected.

You estimate first, then weigh and compare. Say the number out loud or write it down before you put the plate on the scale, otherwise your memory can play tricks on you. Repeat the exercise for a few days, with the foods you eat most often. You will see quite quickly which foods you always get wrong in the same direction [13]. Apps that estimate carbohydrate from photographs are an extra help, not a replacement for the scale [19].

The second check comes from your glucose level. The glucose sensor shows you, two or three hours after the meal, the consequence of your carbohydrate estimate. If the same meal always pushes you above your glucose target, you need to know that carbohydrate is only one of the explanations. The insulin-to-carbohydrate ratio counts too, as do the timing of the dose, the fat in the meal and the glucose level before the meal [20]. Discuss the difference with your doctor, who decides together with you what changes and what stays.