1.

Introduction

Carotenoids form an integral part of the human antioxidant defense network, which protects the body against cellular damage caused by the harmful actions of free radicals.^1–3 Excessive amounts of free radicals can cause oxidative stress, which has been linked to a variety of negative health consequences.^4,5 Carotenoids potentially provide a protective effect against a variety of disorders linked to oxidative stress, including cardiovascular disease, certain cancers, and eye disorders,^2,6–8 although two studies found an increased incidence of lung cancer after beta-carotene supplementation in heavy smokers and individuals exposed to asbestos.^9,10 A protective effect is also consistent with the findings that low carotenoid levels were associated with increased all-cause mortality in a large study of U.S. adults.¹¹ Carotenoids cannot be synthesized by the human body and are obtained from the diet, primarily from red, orange, or yellow pigmented fruit and vegetables.^2,12 Plasma carotenoids are deposited in various human tissues, including the liver, adipose tissue, and skin.^13,14

Carotenoids accumulate in human skin by either diffusing from the blood, through the hypodermis and dermis to the epidermis, or transportation to the epidermis via sweat, which contains carotenoids.^14,15 The two most prominent carotenoids in the skin are beta-carotene and lycopene.¹⁴ Studies have shown that ultraviolet radiation¹⁶ and infrared radiation¹⁷ reduce the concentration of carotenoids in the skin, presumably because the carotenoids are destroyed through their interaction with radiation-induced free radicals. Oxidative stress is associated with enhanced skin aging,¹⁸ and conversely, skin with a high carotenoid concentration appears younger¹⁴ and has fewer furrows and wrinkles than skin with a low carotenoid concentration.¹⁹ Supplementation with beta-carotene (and increased dietary intake of fruit and vegetables) also produces an increase in the normal skin yellowness^20–22—but not skin redness or luminance²¹—of Caucasian skin, and studies have shown that a somewhat yellower skin color is considered healthier and more attractive by both African and Caucasian observers.^20,23–25

Skin carotenoid concentrations are highly correlated with serum carotenoid concentrations^26–28 and can, therefore, provide a noninvasive index of systemic carotenoid concentrations. Skin carotenoid concentrations can also serve as biomarkers for fruit and vegetable consumption,^20–22,29 the antioxidative capacity of human skin, oxidative stress, in general,^30,31 and possibly vitamin A levels [since beta-carotene is a precursor of vitamin A (Ref. 32)]. Another advantage of investigating skin carotenoid concentrations is their use in appearance-based interventions.²² Whitehead et al.³³ found that participants reported a significant, sustained increase in their fruit and vegetable consumption after researchers demonstrated the beneficial effect of fruit and vegetable consumption in a digital manipulation of the participant’s own facial image. Participants who received health information and those who saw the same digital manipulation in a generic face did not report a significant increase in fruit and vegetable consumption.³³

Very little is known about the influence of carotenoid supplementation on African skin. First, although various studies have shown an increase in skin carotenoid concentration and skin yellowness due to carotenoid supplementation in Caucasian skin,^{20,21,27,28,34} to our knowledge, no study has yet examined the effect of carotenoid supplementation on African skin. Second, we previously found that African observers significantly increased skin yellowness (and lightness) in same race facial images to increase the facial images’ apparent health.²⁰ It is, however, still unclear whether moderate increases in carotenoid concentrations produce perceivable yellow color differences in African skin, especially under natural sunlight conditions.

To address these gaps in the literature, this study aims to test whether beta-carotene supplementation produces a significant carotenoid-specific color change in three different regions of African skin: lightly pigmented skin, highly pigmented skin with low sun exposure, and highly pigmented skin with high sun exposure.

2.

Materials and Methods

3.

Results

3.2.

Repeated Measures GLM

Muachly’s test indicated that the assumption of sphericity had been somewhat violated for measurement area [${\mathrm{x}}^2(14)=25.84$, $p=0.049$); therefore, degrees of freedom were corrected using Greenhouse–Geisser estimates of sphericity ($\epsilon =0.42$). The repeated measures GLM analysis of CIELab b* revealed a significant main effect of measurement area [$\mathrm{F}(2,30)=5.530$, $p=0.018$, $\eta p_{\text{partial}}^2=0.480$, $\text{observed power}=0.768$] and week [$\mathrm{F}(8,48)=2.375$, $\mathrm{p}=0.031$, $\eta p_{\text{partial}}^2=0.284$, $\text{observed power}=0.836$]. There was also a significant interaction between week and measurement area [$\mathrm{F}(\mathrm{40,240})=2.235$, $p\le 0.001$, $\eta p_{\text{partial}}^2=0.271$, $\text{observed power}=1.000$]. These results indicate that there was a difference in skin yellowness between the different measurement areas (Fig. 1). More importantly, results show that there was an increase in participants’ skin yellowness over the 8-week supplementation study (CIELab b* before: $\text{mean}=18.856$, $\mathrm{s}.\mathrm{d}.=1.807$; CIELab b* after: $\text{mean}=19.165$, $\mathrm{s}.\mathrm{d}.=2.418$), with certain measurement areas showing a larger change in CIELab b* than others (Fig. 1). The interaction between week and measurement area (but neither of the main effects) was still significant at the conservative Bonferroni adjusted level of $\alpha =0.005$.

Fig. 1

The change in skin yellowness (CIELab b* values) over the 8-week supplementation study for different measurement areas. The palm (solid square) and the inner arm (solid triangle) were the only two measurement areas that showed a significant increase in CIELab b* values.

Separate repeated measures GLMs for each measurement area were used to determine which measurement areas showed significant increases in skin yellowness. When analyzed separately, only the palm [$\mathrm{F}(8,48)=4.324$, $p=0.001$, $\eta p_{\text{partial}}^2=0.419$, $\text{observed power}=0.988$] and inner arm [$\mathrm{F}(8,48)=3.002$, $p=0.008$, $\eta p_{\text{partial}}^2=0.333$, $\text{observed}\text{power}=0.923$] measurement areas showed a significant increase in skin yellowness over the 8-week supplementation study (all other measurement areas $p\ge 0.11$; Fig. 1). The CIELab b* increase in the palm remained significant at Bonferroni adjusted alpha, but the CIELab b* increase in the inner arm did not. Since CIELab a* (skin redness) values in the palm of the hand also increased substantially after supplementation (Table 1) and since basal CIELab b* and CIELab a* are highly correlated in the palm of African skin (${\mathrm{r}}_s=0.786$, $p=0.036$), we performed a repeated measures GLM of CIELAb a* in the palm of the hand to compare the variance explained by CIELab b* and CIELab a*. The palm measurement area also showed a significant increase in skin redness over the 8-week supplementation study [$\mathrm{F}(8,48)=3.419$, $p=0.003$, $\eta p_{\text{partial}}^2=0.363$, $\text{observed power}=0.956$], but CIELab a* explained less variance than CIELAb b* ($\eta p_{\text{partial}}^2=0.419$), indicating that CIELab b* is driving the association between beta-carotene supplementation and skin color change.

3.3.

Change in Spectral Reflectance due to Dietary Supplementation of Beta-Carotene

To verify whether the change in CIELab b* was likely to be due to a change in carotenoid coloration, we calculated the change in spectral reflectance values across the supplementation study for the palm and inner arm (week 8 to week 0). The differences were then averaged across all participants for each position (Fig. 2). Lower values indicate that a smaller proportion of incident light is reflected (which is consistent with a greater light adsorption by skin pigments). Based on the literature, we expected a decrease in spectral reflectance values at 450 to 490 nm—the wavelengths associated with peak light absorption by beta-carotene^20,36—if skin carotenoid coloration increased across the supplementation study. As expected, we observed a decrease in spectral reflectance values at $\sim 450$ to 490 nm (especially for palm measurements), indicating that the change in CIELab b* values were likely to be due to a change in carotenoid coloration (Fig. 2).

Fig. 2

Change in spectral reflectance values due to dietary supplementation of beta-carotene. The graph shows a trough at $\sim 450$ to 490 nm for both measurements (especially the palm), which is consistent with an increase in light adsorption by skin carotenoid pigment.

4.

Discussion

The aim of this study was to determine whether beta-carotene supplementation produces a significant yellow color change in different regions of African skin. To our knowledge, this is the first study to test the effect of carotenoid supplementation in African skin. Our results show that oral supplementation with a daily dose of 15 mg beta-carotene over the course of 8 weeks was associated with a significant increase in skin yellowness in African skin, primarily in lightly pigmented skin (e.g., palm), but also to some extend in highly pigmented skin with low sun exposure (e.g., inner arm). Consistent with previous findings in Caucasian skin,²⁰ beta-carotene supplementation was most strongly associated with CIELab b* skin color changes in the palm measurement area (explaining 42% of the variance). CIELab a* values also increased significantly in the palm measurement area across the supplementation study. This is not unexpected, given the high correlation between CIELab b* and CIELab a* values in African skin.^24,25 Our results show that CIELab b* is, however, driving the association between beta-carotene supplementation and skin color change. Palm measurements of CIELab b* levels might, therefore, serve as a potential marker for systemic carotenoid concentrations;^26,27 fruit and vegetable consumption;^20–22,29 the antioxidative capacity of human skin; oxidative stress, in general;^30,31 and possibly vitamin A levels in the African population [since beta-carotene is a precursor of vitamin A (Ref. 32)].

We should point out some limitations to the study. First, the study had a fairly low sample size. Nevertheless, repeated measures substantially increase the power of the analysis (e.g., Ref. 37), and post hoc power analyses indicate that the study had sufficient power ($\text{observed power}>0.8$) to detect a beta-carotene supplementation effect on skin coloration. Second, we did not include a washout period before and after the supplementation study, which might have increased the noise in beta-carotene levels. Third, we did not measure plasma carotenoid levels directly. It is, however, highly likely that the plasma carotenoid levels increased after supplementation, given that we observed a significant increase in carotenoid-specific coloration in the palm of the hand after beta-carotene supplementation. Nevertheless, future studies could benefit from including plasma carotenoid levels, washout periods, and a larger sample of male and female participants.

We did not find a significant increase in skin yellowness in highly pigmented skin that has high sun exposure (e.g., outer arm and face). This is most likely because an increase in melanin pigmentation due to increased sun exposure in early spring—as evidenced by decreased CIELab L* values (Table 1)—overshadowed the changes in carotenoid pigmentation. Another nonmutually exclusive explanation is that carotenoids accumulated during the supplementation study may have been oxidized through increased ultraviolet and infrared exposure during early spring. It is, therefore, unlikely that participants will be able to observe a beneficial effect of fruit and vegetable consumption in sun-exposed highly pigmented African skin under natural sunlight conditions. Of course, a change in carotenoid coloration might also be observable in sun-exposed highly pigmented African skin during late summer/early winter or if participants consume more fruit and vegetables or higher doses of carotenoid supplementation.

In summary, we found a significant increase in skin yellowness of African skin after an 8-week beta-carotene supplementation study, but only in lightly pigmented skin (e.g., palm) and to some extend in highly pigmented skin with low sun exposure (e.g., inner arm). Skin carotenoid measurements in the palm of the hand or sun-protected skin areas might, therefore, serve as a potential marker for systemic carotenoid concentrations in people of African descent.