Removing an E-number from your sauce is not one decision, it is two, and buyers keep collapsing them. The first is a process question: can a non-chemical starch route absorb the acid, heat, shear, freeze-thaw and shelf-life stress the grade currently carries? The second is a naming question: does the ingredient you switch to let you drop the word “modified” or the additive code on the label? They have opposite failure modes. A physically modified or annealed starch may satisfy the naming question yet fail the process question and thin out in the retort; another chemical grade may solve the process question and still print an E-number. Before you change anything, run each candidate through both gates separately, and expect the honest answer to be “cleaner label is achievable only for some functions.”
What Does “Cleaner Label” Actually Mean to a Food Manufacturer?
To a consumer, clean label is a feeling: fewer words, familiar ingredients. To a manufacturer it has two concrete edges, and both have to be checked.
The first edge is naming. Some starches declare as “modified starch” or carry an E-number, while others declare as a source name (tapioca starch, potato starch) or simply as “starch.” The second edge is function: the reason the chemical grade went into the recipe in the first place, usually tolerance to a processing stress that a native starch does not survive.
“Clean label” is not a single regulated claim. It is a marketing umbrella. What a supplier can document is objective — ingredient identity, base starch, allergen status, how the starch was treated. What no supplier can document is the word “clean” itself, because there is no fixed legal definition it has to meet. That is why the reformulation decision belongs to you, not to a grade name.
Which Processing Stresses Can a Non-Chemical Route Absorb?
Chemical modification buys you specific things: a cross-link band that resists shear and acid, a substitution that lowers gelatinisation temperature or suppresses retrogradation. A native, annealed or physically modified starch usually keeps only some of that behaviour, and which parts survive depends on the stress.
Think of the stresses as a list you test one at a time:
- Acid at temperature. Low pH plus heat hydrolyses starch chains over time. Cross-linked grades slow the loss; native and physically modified pastes thin measurably in an acid hold. Hold a sample at your finished pH and your process temperature for your process time, and read viscosity against time.
- Retort or UHT heat load. The thermal death time your product needs is often beyond what an unmodified granule network survives. Run the real thermal process on filled packs and measure post-process viscosity and texture, not just lab-cooked paste.
- Shear. Homogenisers, high-shear mixers, pumps and colloid mills tear swollen granules. Physically modified and native starches break down earlier on an ascending flow curve. Compare viscosity before and after the suspect unit operation.
- Freeze-thaw. Syneresis on thaw is where native starches fail hardest; retrogradation and weak granule structure release water. Cycle packs through several freeze-thaw cycles and weigh the serum.
- Long ambient shelf life. The setback and clarity a grade holds over months is a different property from its cook behaviour. Do not read a single peak value as a shelf-life result.
The point of the list is that “can I remove it” is really five questions, and a clean route usually passes some and fails others. A sauce that only needs body and gloss in a mild, hot-fill process may pass all five on a physical route. A low-pH retort dressing usually will not.
Which Starch Names Can Drop “Modified” or an E-Number?
Once you know which stresses you can live without, the naming gate decides what the switch is actually worth on the label. Naming is set by regulation, and it differs by market, so treat it separately from performance.
In the United States, 21 CFR 172.892 defines “food starch-modified” by the treatments it allows: acid modification, bleaching, oxidation, esterification and etherification, with defined limits, plus certain enzyme and alkali steps. A starch modified only by heat, such as annealing or heat-moisture conversion, is not named by that list, so heat-only treatment does not by itself force the “modified” declaration. Native starches still declare by their source name. Where the treatment is unusual, confirm the declaration with a regulatory adviser rather than assuming.
In the European Union, starches treated only physically, enzymatically, or by acid, alkali or bleaching are generally treated as ingredients rather than as food additives, so they are commonly listed without an E-number, while chemically cross-linked or substituted starches are additives and are reported to carry numbers in the E1404–E1452 range. The label wording, “starch” versus “modified starch,” follows the treatment actually used. Because additive status and naming are matters of EU law rather than of supplier preference, confirm the exact route with a regulatory adviser before you change a label.
The asymmetry that surprises buyers is this: a grade can solve the naming gate and still fail the process gate, or the reverse. Dropping the E-number is not the same as dropping the function, and in a hard process it often cannot be had for free.
Can Dextrin or Maltodextrin Replace a Modified Starch?
Only for part of the job, and buyers often pick them for the wrong reason. Dextrins and maltodextrin are usually declared as ordinary ingredients rather than as additives, so they look clean on paper. But they are not thickener networks. They contribute body, solids, bulk and mouthfeel; they do not build the acid-and-shear-resistant paste a cross-linked starch builds.
So the honest use of a dextrin in a clean-label reformulation is as a filler or body partner, not as a one-for-one stand-in. If you swap a modified starch for maltodextrin at equal weight, you lose viscosity and probably gain sweetness and solids. Read the DE value as a functional dial, not as a label trick, and pair the dextrin with a starch route that carries the actual thickening.
Is Pregelatinised or Annealed Starch “Cleaner”?
Pregelatinised starch is physically modified: it is cooked, gelatinised and dried so it thickens in cold water. Annealing is a heat-moisture treatment that changes granule behaviour without adding chemistry. Both are attractive on the naming gate, because in the US a heat-only route is not what 172.892 calls modified, and in the EU a physically modified starch is generally not an additive.
The catch is function. A cold-water-swelling pregelatinised paste is built for convenience and cold dispersion, not for surviving a retort hold or a low-pH shelf. It answers “can I declare this as starch,” not “can this replace my acid-heat grade.” The choice between a physical and a chemical route is a choice of which stresses your line actually applies. For the mechanism behind that difference, see pregelatinised versus chemically modified starch.
How Do I Test a Clean Route Without Copying Someone’s Numbers?
Do not transplant published viscosity figures from a paper or a competitor sheet into your product. Those values were measured on a different matrix, at a different pH, with a different shear and thermal history. Build the comparison on your own line and your own lab, in this order.
First capture your baseline on the current grade: viscosity at fill, after the heat treatment, after cooling, and after a representative shelf interval, plus serum on freeze-thaw if that applies. Second, set an acceptance band from that baseline for the checkpoints your process actually depends on. Third, run the clean candidate through an RVA curve with the same method for both materials, so the curves are comparable; the counter-match logic is covered in the starch counter-match guide. Fourth, where the curve looks close, do not stop there — the curve is a screen, and a passing curve still has to survive the acid hold, the real thermal process and the shear station. Only at that point do you commit to a label change.
Where the native or physical base genuinely cannot carry a stress, the honest options are a blend of a clean route plus a small amount of a hydrocolloid, a change to the process window, or keeping the modified grade and accepting the label. Deciding which is a cost conversation about reformulation and re-qualification, not about ingredient names.
FAQ
Does removing E1442 automatically make my label cleaner?
No. Removal is a naming result, and only helps if the substitute both declares without an additive code and survives your process. Those are two independent checks, and a substitute can pass one and fail the other.
Is a dextrin cleaner than a modified starch?
It is usually declared as an ordinary ingredient rather than an additive, so it reads cleaner. But it is not a thickener network, so it replaces body, not acid-and-shear stability. Treat it as a partner, not a stand-in.
Will annealing let me drop the “modified” name?
In the US, heat-only treatments are not the treatments 172.892 uses to define “food starch-modified,” so annealing alone does not force that name; native source names still apply. In the EU, confirm the route with a regulatory adviser before you change the label.
What can a starch supplier actually document for a clean-label claim?
Objective attributes: identity, base starch, treatment used, allergen and gluten status, and test results. They cannot document the marketing word “clean,” and you remain responsible for any consumer-facing claim on the finished pack.
Can one clean blend replace a modified starch in a retort dressing?
Usually not at no cost. A retort plus low pH is the hardest case, and a single physical or native starch rarely carries both. Expect a blend, a narrower process window, or the modified grade.
If you are deciding which stresses your line can live without, send GFMStech your process window and target label wording, and we will build a clean-label starch solution around what your process actually needs to keep, using RVA counter-match data rather than a code swap alone.
