Colour as a Health Cue

What it is

The idea is that shoppers do not read a pack the way a nutritionist would. They take a fast impression from what is easiest to see, and colour is among the first things seen. If a colour reliably comes to stand for health, a brand can borrow the meaning without making a claim.

The claim rests on four separate lines of evidence, and they should not be merged. The first is Jonathon Schuldt's 2013 pair of experiments in Health Communication. In the first, 93 university students imagined being hungry in a grocery checkout line and saw a candy bar with either a red or a green calorie label. The calorie count was identical, and the bar under the green label was rated healthier. In the second, 39 online participants saw candy with a green or a white label and rated how much healthy eating mattered to their food choices. The more a person valued healthy eating, the less healthful the white-labelled bar looked, and the green label removed that pattern. Those sample sizes come from Cornell University's description of the paper at the time. What was coloured is a label, a piece of information, and not the pack.

The second line is the 2023 review by Katharina Steiner and Arnd Florack in Foods, which pooled 20 articles containing 35 studies. Cool colours, blue and green, were linked to higher healthiness ratings in most studies though not all. Green labels often helped people identify healthier products. The advantage was not consistent: in one study it held for men only, and in another a white label with a smiley face beat both green and red labels. The authors call green a candidate for a colour that raises health judgements across contexts, and add that where health is unexpected, as with chips, the effect might disappear.

The third is saturation, the vividness of a colour. James Mead and Rob Richerson (2018) ran four experiments, with potato chips and nutrition bars among the products according to the review, and found that foods in highly saturated packaging were rated less healthful than foods in muted packaging. They attribute this to conceptual fluency, since shoppers meet unhealthy foods in saturated packaging often. Subjective nutrition knowledge weakened the effect and restrained eating strengthened it. Sonja Kunz, Simona Haasova and Arnd Florack (2020) found the reverse with drinks. In their second study, 171 participants rated the same 20 drinks in versions made 25 per cent more and 25 per cent less saturated, with other colour dimensions held constant. The more saturated versions were rated healthier and tastier, and perceived freshness explained the health effect completely. The health difference was very small.

The fourth is the other meaning of green. Ethan Pancer and colleagues (2017) found in three experiments that a green colour without an eco-label, or an eco-label without green, lowered perceived product efficacy. That is a result about how well a product is expected to work, not about health, and it shows that green is read through more than one idea.

In effect

For a brand, the practical reading is that a colour should be tested in its own category before it is expected to carry a health message. The same green sits differently on a juice and on a snack bar, and the review lists product category, country, gender and attitudes to healthy eating as moderators.

For a shopper, the cue is easy to over-read. Colour is designed, while the nutrition panel is a measurement. A pack that looks healthy has told the viewer how its maker wants it read.

The review also proposes three routes from colour to judgement, and says they can operate together: colour read as an explicit health signal, colour that triggers beliefs about the food, and colour that prompts a mental simulation of eating it. These are a framework for explaining the pattern. The authors present them as something for future work to test.

What it does not say

It does not say that green makes any food seem healthy. The review notes that most studies relied on self-report and that methods varied widely, and its authors hedge about green in categories such as chips.

It does not say that vivid colour is a health signal or a warning. Two studies on saturation point in opposite directions, and the gap between them can be reconciled on paper by product category. That reconciliation is a hypothesis about the gap and not a result.

It does not say that a rating changes behaviour. The studies described here collected ratings of labels or images. The outcome is what people said about a food, not what they bought or ate.

It does not come from the social-priming literature, and the distinction matters. Adam Alter's Drunk Tank Pink places colour among the cues that steer behaviour without notice, but the vault's audit of that book flags that much of it rests on priming research that has not replicated. The packaging studies ask for a judgement and measure the judgement.

---