Two sensory studies illustrate why bitterness masking needs to be evaluated in the intended beverage after processing. They cover American-ginseng low-lactose milk and separate ginseng model solutions—not interchangeable ingredients or formulations. The trial framework below is proposed, not a validated recipe or a report of factory trials.
Ask where bitterness changes before choosing a masking agent
A 2012 study of low-lactose milk containing American ginseng (Panax quinquefolius L.) provides a useful warning. After UHT processing, the beverage contained 7.52 mg total ginsenosides per 100 g milk, corresponding to 67.6% recovery relative to the unprocessed extract under the study’s comparison. Yet the industrially processed ginseng milk had the strongest brown color, bitterness and metallic taste.[1] That does not prove that a decrease in total ginsenosides caused stronger bitterness. It shows that analytical recovery and sensory performance cannot substitute for one another in that system.[1]
Link ingredient selection, heat processing and flavor screening through the same set of samples: record ingredient addition and ginsenoside composition alongside bitterness, metallic taste and aftertaste in the processed beverage.
What the direct sensory evidence establishes
The milk study analyzed individual ginsenosides by HPLC and used a trained panel to assess flavored and unflavored milks. Its abstract reports reduced ginseng-related bitterness after the addition of vanilla flavor and sucralose.[1]
The abstract provides observations on post-process composition and flavor, but not the complete thermal history, sensory scale or masking recipe. The reported 67.6% recovery is relative to the unprocessed extract; interpreting the unrecovered fraction requires the analytical recovery methods.[1]
Why a separate “bitterness halved” result is not a milk recipe
A 2010 study evaluated masking treatments in water and a model energy-drink base. Its abstract describes 0.052 g of a panax ginseng material labeled as 80% ginsenosides in each 100 mL of solution. Twelve trained panelists assessed 42 treatments, with and without nose clips.[2] In that screening, 0.09 g γ-cyclodextrin per 100 mL or 1 g β-cyclodextrin per 100 mL each reduced bitterness intensity by half.[2]
These amounts are study conditions, not recommended milk-beverage use levels. Trained-panel bitterness scores also differ from consumer liking.[2] Cyclodextrins are candidates for screening in the intended beverage, alongside vanilla and sweetener systems, with post-process bitterness, marker retention and storage performance assessed separately. Reassess these outcomes when changing ginseng species or beverage base.
How heat processing changes ginsenoside composition
An openly available hydrothermal-processing paper did measure ginsenoside transformation. It used Panax ginseng extract adjusted to 4 wt% solids, 23 mL batch reactors and reaction temperatures of 100–160°C, with HPLC and chemical antioxidant endpoints.[4] The authors discuss competing formation and decomposition rates; across the reaction times tested at each temperature, the highest observed 20(S)-Rg3 yield decreased as temperature rose, whereas the highest Rk1 and Rg5 yields increased; those maxima did not all occur at the same reaction time.[4]
That was not a milk UHT experiment and it did not report human bitterness ratings. Its chromatographic changes cannot establish a causal chain of “heat → this particular ginsenoside → stronger bitterness” in the milk beverage.[4] A practical investigation should record pre/post-process composition and sensory changes side by side, leaving attribution to an individual compound as a question to test.
A trial structure that separates possible causes
Use the comparisons below to plan trials. Follow the project’s validated food-safety process and taste only samples cleared as safe; do not reduce thermal severity to improve flavor or assume an untreated laboratory control is safe to taste.
Scroll horizontally to read all three columns; the comparison label stays visible.
| Comparison | Conditions to align | Question it addresses |
|---|---|---|
| Base without ginseng, sampled before/after the target process | Same milk or plant base, sweetener system and other ingredients | How much flavor or color change comes from processing the base itself? |
| Base with ginseng, sampled before/after the target process | Same ingredient batch and addition rate, same thermal process | Does the ginseng-containing system change beyond the base response? |
| Processed ginseng product, without and with candidate masking systems | Actual addition sequence, processing and packaging conditions | Does the candidate still work in the final product? |
| Masking-system blank without ginseng | Same candidate level and processing | Does the candidate add sweetness, aftertaste or another sensory effect of its own? |
State the comparison basis in advance. The same ingredient batch at the same addition rate suits a pre/post-process comparison. A supplier comparison also needs botanical species, processing identity, assayed composition and carriers. Do not quietly change total solids to force equal total ginsenosides and then attribute every sensory difference to the extract.
The proposed records have three parallel tracks:
- Analytical: Define the total, the individual peaks included, and the extraction and recovery method. Where settling or container adhesion occurs, retain a representative whole-container sample and the abnormal fractions rather than comparing only clear supernatants without explanation.
- Sensory: Score bitterness, metallic taste, ginseng character, sweetness and aftertaste separately. Randomize sample codes and control serving temperature. Keep trained descriptive analysis distinct from consumer liking.
- Process: Record actual thermal history, addition sequence, color, physical condition, packaging and planned storage points. Distinguish independent preparation batches from repeated measurements of one bottle.
Set analytical and sensory acceptance criteria before screening, then check them in processed samples; an acceptable cold prototype does not replace this step.
Include processing requirements in material selection
Replace a generic request for a “high-ginsenoside extract” with a one-page material and application brief: species and plant part, processing identity such as white or red ginseng where applicable, target base, thermal process, ingredient mass per serving, unacceptable sensory attributes and analytical definitions. Ask for traceability between the trial sample and subsequent supply in composition, carriers and processing identity—not just one total-ginsenoside figure.
Evaluate post-process flavor and total-ginsenoside recovery separately. The American-ginseng milk and ginseng model-solution studies identify candidates for flavor adjustment; trials in the intended beverage establish how those candidates perform with the actual ingredient, base and process.[1][2]
Download the blank trial record
Download the English record (.txt)
The record leaves observations and acceptance criteria blank; it is a planning aid, not a validated SOP. For material and process distinctions, read the ginseng ingredient guide; discuss sample identity through ginseng sourcing. The wider beverage application context helps frame the intended base and processing task.
Sources
[1] Active compounds and distinctive sensory features provided by American ginseng (Panax quinquefolius L.) extract in a new functional milk beverage. Original abstract; full methods and sensory tables not available in this review.
[2] Sensory properties of ginseng solutions modified by masking agents. Original abstract; full methods and sensory tables not available in this review.
[4] Effect of hydrothermal processing on ginseng extract — full text Full-text XML, methods and Results 3.1; no human bitterness endpoint.
Related Products
Interested in our botanical extracts?
Contact us for samples, technical documentation, or custom formulation support.
Get in Touch