The authors have declared that no competing interests exist.
Red hair is associated in women with pain sensitivity. This medical condition, and perhaps others, seems facilitated by the combination of being red-haired and female. We tested this hypothesis by questioning a large sample of Czech and Slovak respondents about the natural redness and darkness of their hair, their natural eye color, their physical and mental health (24 categories), and other personal attributes (height, weight, number of children, lifelong number of sexual partners, frequency of smoking). Red-haired women did worse than other women in ten health categories and better in only three, being particularly prone to colorectal, cervical, uterine, and ovarian cancer. Red-haired men showed a balanced pattern, doing better than other men in three health categories and worse in three. Number of children was the only category where both male and female redheads did better than other respondents. We also confirmed earlier findings that red hair is naturally more frequent in women than in men. Of the ‘new’ hair and eye colors, red hair diverges the most from the ancestral state of black hair and brown eyes, being the most sexually dimorphic variant not only in population frequency but also in health status. This divergent health status may have one or more causes: direct effects of red hair pigments (pheomelanins) or their by-products; effects of other genes that show linkage with genes involved in pheomelanin production; excessive prenatal exposure to estrogen (which facilitates expression of red hair during fetal development and which, at high levels, may cause health problems later in life); evolutionary recentness of red hair and corresponding lack of time to correct negative side effects; or genetic incompatibilities associated with the allele Val92Met, which seems to be of Neanderthal origin and is one of the alleles that can cause red hair.
It has long been known that redheads are at higher risk of sunburn and skin cancer. This is to be expected because red hair is associated with fair skin, which is more vulnerable to UV radiation [
The risk factor seems to be specific to women. Pain sensitivity is higher in female redheads than in male redheads [
Prenatal estrogen may therefore mediate the relationship between red hair and certain aspects of health, including some that remain unsuspected. It was only by chance that researchers discovered the three-way association between being a woman, having red hair, and feeling more sensitivity to pain. There has been no systematic effort to identify all female-specific associations between human health and red hair, let alone between human health and each of the different hair and eye colors.
For these reasons, we wished to find out how different aspects of human health vary as a function of hair/eye color. We also wished to see how well the variance is explained by the two known risk factors: 1) vulnerability to UV, as measured by relative importance of skin cancer; and 2) gender, specifically being a woman. To this end, we used an existing dataset collected for an unrelated purpose: a survey on the RhD factor in relation to various health categories in a Czech and Slovak population. This survey encompassed a very large number of individuals and could thus capture relatively weak associations between health status and other factors.
The present study reanalyzed data originally collected for a survey on the RhD factor in relation to human health. Respondents were recruited by a Facebook-based snowball method [
All participants provided informed consent by pressing the corresponding button on the electronic form. All methods were performed in accordance with the relevant guidelines and regulations. The study, including the method of obtaining informed consent (by pressing the Next button on the first page), was approved by the IRB of the Faculty of Science, No. 2014/21."
The questionnaire was distributed as a Czech/English Qualtrics survey (
The medical part of the questionnaire was prepared by two physicians: a clinician (internist/hematologist), and a researcher (molecular geneticist). Questions fell into two parts, one using subjective measures of health status and the other more objective measures. Respondents were first asked to rate the presence and intensity of their health problems on a 6-point Likert scale. These questions were on physical health and mental health in general and on more specific health categories: allergies; cancer; digestion; fertility; genitourinary system; heart and vascular system; hematology; immune system; metabolism, including endocrine system; musculoskeletal system; nervous system; respiratory organs; sense organs; and sexual function. The second part of the questionnaire was designed to provide objective information on health status. For example, respondents were asked how many physician-prescribed drugs they were currently taking per day, how many “different herbs, food supplements, multivitamins, superfoods etc.” they were currently taking per day, and how often they had used antibiotics during the past 365 days.
As a benchmark for the relative strength of associations between hair/eye color and the 24 health categories, we measured the associations between these categories and two unrelated but well-known risk factors: body mass index (BMI) and smoking.
For some of these categories, we also asked the respondents to state the specific disorders they had or used to have. For the ‘Cancer’ category, respondents were asked “What kind of cancer are you suffering from or have you suffered from?” They then read a list of disorders and ticked the appropriate boxes.
Statistica v. 10 and IBM SPSS v. 21 were used for most of the statistical analysis. MANCOVAs (with gender, eye color, or hair color as predictors) were performed by the “adonis” function available within Vegan package in R [
Information on hair/eye color was provided by 2,558 men and 4,472 women out of 7,044 Czech and Slovak respondents (the others did not complete the questionnaire part of the test). Mean age was higher for the men (36.8, std. dev. 13.5) than for the women (34.6, std. dev. 13.0) t7028 = 6.9, p < 0.0005.
Figs
Respondents rated hair redness on a scale of 1 to 6 where 1 = not at all red and 6 = completely red.
Age was associated in women with darker eyes, darker hair, and redder hair (
Hair darkness | Hair redness | Eye darkness | Age (men) | Age (women) | |
---|---|---|---|---|---|
Hair darkness | 0.00 | ||||
Hair redness | -0.01 | ||||
Eye darkness | 0.00 | 0.02 |
The upper right of the table (excluding the last two columns) shows partial Kendall Tau correlations (age controlled) for men, and the lower left the same results for women. The last two columns show standard Kendall Tau correlations between hair/eye color and age in men and women, respectively. Significant correlations are in bold.
We looked for significant associations between 24 health categories and different hair/eye colors. Yellow eye color was reported by only 5 respondents and therefore excluded from the analyses. The results, shown in
Blue | Green | Brown | Black | Grey | Amber | Hazel | Eye darkness | Hair darkness | Hair redness | BMI | Smoking | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Physical health problems in general | -0.015 | -0.019 | 0.011 | 0.004 | 0.015 | |||||||
Mental health problems in general | 0.009 | 0.014 | 0.014 | 0.015 | 0.003 | 0.011 | -0.024 | |||||
Specific health problems: | ||||||||||||
Antibiotics/ |
0.007 | -0.001 | -0.004 | 0.002 | -0.018 | -0.002 | 0.006 | 0.014 | ||||
Acute care/ |
0.008 | -0.006 | -0.004 | -0.007 | 0.026 | 0.014 | -0.008 | 0.011 | ||||
Different med. |
0.002 | -0.003 | 0.002 | 0.026 | 0.002 | -0.015 | -0.012 | -0.001 | -0.021 | 0.017 | ||
Number of |
0.015 | 0.023 | -0.014 | -0.017 | -0.012 | -0.016 | -0.026 | |||||
Number of |
0.021 | 0.000 | -0.019 | -0.020 | 0.001 | -0.015 | 0.007 | -0.003 | -0.010 | 0.013 | ||
Allergies | -0.001 | 0.003 | -0.004 | -0.014 | 0.015 | 0.020 | -0.012 | 0.015 | -0.001 | |||
Immunological | 0.019 | -0.003 | 0.023 | -0.023 | -0.005 | -0.003 | 0.000 | -0.015 | ||||
Digestive | 0.018 | 0.004 | 0.025 | 0.001 | 0.020 | -0.017 | 0.012 | |||||
Heart & |
-0.022 | 0.012 | -0.007 | 0.003 | 0.003 | 0.005 | -0.008 | |||||
Hematological | 0.018 | 0.018 | -0.015 | -0.023 | 0.003 | -0.011 | -0.028 | -0.017 | 0.024 | -0.002 | -0.009 | |
Metabolic | 0.018 | -0.013 | -0.001 | 0.013 | 0.012 | -0.009 | -0.007 | -0.024 | 0.022 | -0.017 | ||
Cancer | 0.013 | -0.030 | -0.015 | -0.010 | -0.004 | 0.000 | -0.020 | 0.018 | ||||
Fertility | -0.015 | -0.023 | 0.012 | 0.012 | 0.006 | 0.000 | -0.003 | 0.017 | 0.001 | |||
Genitourinary | 0.016 | -0.008 | -0.004 | -0.024 | -0.009 | 0.001 | 0.013 | -0.013 | -0.025 | 0.018 | -0.011 | -0.028 |
Sense organs | -0.002 | 0.025 | -0.005 | -0.004 | -0.018 | 0.023 | 0.005 | 0.004 | 0.017 | 0.009 | ||
Neurological | -0.023 | 0.007 | 0.003 | -0.012 | -0.001 | 0.008 | 0.005 | 0.023 | -0.018 | -0.012 | ||
Psychiatric | 0.001 | 0.014 | 0.007 | 0.002 | 0.027 | 0.008 | 0.008 | -0.009 | ||||
Sexual |
-0.003 | 0.003 | -0.011 | 0.020 | -0.006 | -0.022 | -0.011 | 0.007 | ||||
Musculoskeletal | -0.025 | 0.014 | 0.016 | 0.010 | -0.017 | 0.001 | 0.026 | |||||
Respiratory | -0.011 | -0.013 | 0.030 | 0.019 | -0.006 | 0.010 | 0.014 | 0.006 | 0.001 | |||
Tiredness |
0.024 | 0.027 | -0.009 | -0.008 | 0.025 | 0.009 | -0.011 | 0.008 | ||||
Headaches |
-0.009 | 0.023 | 0.020 | 0.017 | -0.002 | 0.010 | 0.010 | -0.011 | 0.008 | |||
Reproductive/ mating success: | ||||||||||||
Number of |
-0.007 | -0.026 | 0.018 | 0.014 | -0.006 | -0.025 | -0.007 | |||||
Number of |
0.018 | -0.002 | 0.020 | 0.025 | 0.000 | -0.005 | -0.003 | 0.009 | ||||
Negative divergences in health status | 1 | 0 | 2 | 6 | 9 | 3 | 3 | 1 | 0 | 3 | 10 | 10 |
Positive divergences in health status | 9 | 2 | 0 | 1 | 1 | 2 | 13 | 0 | 0 | 3 | 3 | 10 |
The figures (age-controlled partial Kendall Tau correlations) show the strength and direction of associations between variables on the top and on the left. A positive figure means a positive association between a respondent characteristic (column headings) and a category of human health, including number of children and sexual partners (row headings). Associations that remain significant after correction for multiple testing are in bold. The last two rows show the total number of significant associations where the divergence in health status is either negative or positive. A higher number of children and a higher number of sexual partners are classified as positive divergences in health status.
Blue | Green | Brown | Black | Grey | Amber | Hazel | Eye darkness | Hair darkness | Hair redness | BMI | Smoking | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Physical health problems in general | -0.007 | 0.010 | -0.020 | -0.009 | 0.011 | |||||||
Mental health problems in general | 0.014 | 0.013 | 0.001 | -0.007 | 0.003 | 0.013 | ||||||
Specific health |
||||||||||||
Antibiotics/ |
-0.005 | 0.012 | 0.015 | -0.005 | -0.001 | 0.002 | 0.013 | |||||
Acute care/ |
-0.013 | 0.015 | 0.010 | -0.007 | -0.009 | 0.002 | ||||||
Different med. |
-0.018 | -0.014 | 0.006 | -0.005 | -0.022 | -0.014 | 0.014 | -0.006 | 0.013 | |||
Number of |
-0.009 | 0.002 | 0.008 | 0.005 | ||||||||
Number of |
0.001 | 0.012 | 0.003 | 0.008 | -0.012 | -0.022 | 0.001 | 0.007 | 0.004 | 0.013 | -0.019 | |
Allergies | -0.001 | 0.011 | -0.011 | -0.001 | -0.002 | 0.016 | -0.012 | |||||
Immunological | -0.015 | 0.019 | 0.006 | 0.007 | ||||||||
Digestive | -0.015 | 0.018 | 0.016 | 0.012 | -0.002 | 0.010 | 0.014 | -0.001 | 0.008 | 0.009 | ||
Heart & |
-0.010 | 0.009 | 0.005 | -0.010 | 0.011 | -0.006 | 0.002 | 0.015 | ||||
Hematological | 0.004 | -0.016 | 0.000 | 0.008 | -0.014 | 0.009 | -0.004 | 0.000 | 0.007 | -0.006 | -0.007 | |
Metabolic | 0.019 | 0.010 | 0.021 | 0.014 | 0.007 | |||||||
Cancer | 0.012 | -0.014 | -0.012 | 0.017 | -0.010 | 0.003 | -0.009 | 0.011 | 0.013 | |||
Fertility | 0.005 | 0.015 | -0.005 | 0.013 | 0.007 | 0.001 | 0.013 | 0.015 | 0.006 | -0.010 | ||
Genitourinary | -0.017 | -0.007 | -0.006 | -0.024 | 0.016 | 0.009 | 0.013 | |||||
Sense organs | -0.002 | -0.007 | -0.016 | 0.005 | 0.003 | 0.000 | 0.011 | |||||
Neurological | -0.013 | -0.003 | 0.006 | -0.001 | -0.002 | 0.011 | 0.014 | 0.011 | -0.005 | |||
Psychiatric | -0.012 | 0.004 | 0.005 | -0.006 | 0.006 | 0.007 | ||||||
Sexual function | -0.012 | 0.003 | 0.012 | -0.016 | 0.016 | 0.006 | -0.002 | 0.015 | 0.017 | |||
Musculoskeletal | 0.000 | 0.010 | -0.005 | -0.007 | 0.005 | -0.015 | -0.005 | -0.004 | -0.018 | 0.013 | ||
Respiratory | 0.006 | -0.006 | -0.008 | -0.008 | 0.013 | -0.006 | ||||||
Tiredness |
-0.004 | 0.005 | 0.001 | 0.003 | 0.016 | -0.007 | -0.005 | 0.014 | 0.014 | |||
Headaches |
-0.001 | 0.006 | -0.004 | 0.013 | 0.010 | -0.009 | -0.009 | 0.015 | -0.002 | 0.014 | ||
Reproductive/ |
||||||||||||
Number of |
0.006 | 0.004 | -0.008 | -0.016 | -0.010 | |||||||
Number of |
0.010 | 0.016 | -0.006 | 0.014 | 0.014 | -0.008 | -0.001 | 0.011 | ||||
Negative divergences in health status | 0 | 2 | 10 | 4 | 7 | 0 | 1 | 7 | 7 | 10 | 14 | 12 |
Positive divergences in health status | 9 | 4 | 2 | 3 | 2 | 0 | 8 | 2 | 1 | 3 | 3 | 3 |
See
To determine the relative importance of skin cancer in the Cancer category, we looked at the incidences of specific types of cancer. The results are shown in
Type of cancer | Men | Women | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Red- Can- | Red-Can+ | Red+ Can- | Red+ Can+ | OR | p | Red-Can- | Red-Can+ | Red+ Can- | Red+ Can+ | OR | p | |
Esophageal cancer | 1762 | 0 | 268 | 0 | 2946 | 1 | 836 | 1 | 3.52 | 0.394 | ||
Stomach cancer | 1762 | 0 | 268 | 0 | 2946 | 1 | 836 | 1 | 3.52 | 0.394 | ||
Colorectal cancer | 1762 | 0 | 264 | 4 | 273.56 | 2947 | 0 | 834 | 3 | 109.53 | ||
Liver cancer | 1762 | 0 | 267 | 1 | 72.57 | 0.132 | 2947 | 0 | 836 | 1 | 38.77 | 0.221 |
Lung cancer | 1761 | 1 | 267 | 1 | 6.59 | 0.247 | 2947 | 0 | 837 | 0 | ||
Melanoma, other skin cancers | 1753 | 9 | 266 | 2 | 1.52 | 0.271 | 2937 | 10 | 833 | 4 | 1.43 | 0.243 |
Breast cancer | 1762 | 0 | 268 | 0 | 2932 | 15 | 830 | 7 | 1.66 | 0.318 | ||
Cervical uterine precancerosis | 1677 | 0 | 259 | 0 | 2723 | 64 | 756 | 27 | 1.52 | |||
Cervical uterine cancer | 1762 | 0 | 268 | 0 | 2902 | 45 | 817 | 20 | 1.58 | |||
Corpus uteri cancer | 1762 | 0 | 268 | 0 | 2940 | 7 | 835 | 2 | 1.04 | 1.000 | ||
Ovarian cancer | 1762 | 0 | 268 | 0 | 2942 | 5 | 832 | 5 | 3.54 | |||
Prostate cancer | 1753 | 9 | 266 | 2 | 1.52 | 0.184 | 2947 | 0 | 837 | 0 | ||
Lymphoma, myeloma multiple | 1760 | 2 | 268 | 0 | 0.31 | 1.000 | 2947 | 0 | 837 | 0 | ||
Leukemia | 1758 | 4 | 268 | 0 | 0.16 | 1.000 | 2941 | 6 | 836 | 1 | 0.63 | 1.000 |
Bladder cancer | 1761 | 1 | 268 | 0 | 0.60 | 1.000 | 2945 | 2 | 836 | 1 | 1.85 | 0.528 |
Mouth, oropharynx cancers | 1761 | 1 | 268 | 0 | 0.60 | 1.000 | 2947 | 0 | 836 | 1 | 38.77 | 0.221 |
Adeno-carcinoma | 1762 | 0 | 268 | 0 | 2943 | 4 | 837 | 0 | 0.09 | 0.582 | ||
Papilloma cancer | 1762 | 0 | 268 | 0 | 2941 | 6 | 833 | 4 | 2.37 | 0.422 | ||
Other types of cancer | 1747 | 15 | 265 | 3 | 1.35 | 0.642 | 2929 | 18 | 825 | 12 | 2.37 |
Red- Can- = Number of non-redheads (i.e., respondents whose intensity of redness is 1–3 on a 6-point Likert scale) without the specific type of cancer
Red- Can+ = Number of non-redheads with the specific type of cancer
Red+ Can- = Number of redheads (i.e., respondents whose intensity of redness is 4–6) without the specific type of cancer
Red+ Can+ = Number of redheads with the specific type of cancer
Odds Ratios (OR) and statistical significance (p) respectively are shown for men and women and for each specific type of cancer. Age was controlled by performing a partial Kendall Tau’s correlation whenever the incidence of a specific type of cancer exceeded 9. Otherwise, a Fisher’s exact test was performed to determine statistical significance. ORs higher than 1 indicate that redness is positively associated with the incidence of the specific type of cancer. Results are in bold if significant in two-sided tests after Benjamini-Hochberg correction for multiple tests, and p-values < 0.0005 are coded as 0.000.
Because divergences in health status differed between men and women, particularly among red-haired respondents, we looked for a significant interaction with gender. To this end, we first performed four MANCOVAs to see whether variance in respondent health correlated significantly with gender, eye darkness, hair darkness, and hair redness. We then performed three MANCOVAs to see whether variance in respondent health correlated significantly with an interaction between gender and any of the other variables: eye darkness, hair darkness, or hair redness. Finally, we constructed a new binary variable—presence or absence of green eyes—and performed two MANCOVAs to see whether variance in respondent health correlated significantly with this new variable or with an interaction between it and gender. All MANCOVAs had respondent age as the covariate. The results are shown in
Independent variable | R | p-value |
---|---|---|
gender | 0.12 | 0.001*** |
hair redness | 0.04 | 0.001*** |
hair darkness | 0.02 | 0.35 |
eye darkness | 0.03 | 0.002** |
gender*hair redness | 0.02 | 0.182 |
gender*hair darkness | 0.01 | 0.583 |
gender*eye darkness | 0.02 | 0.405 |
green eyes vs. all other eye colors | 0.02 | 0.049* |
gender*green eyes vs. all other eye colors | 0.12 | 0.001*** |
This finding made us take a second look at the relationship between female respondent health and hair redness. That relationship, too, might not be fully understood through a linear regression. We specifically looked at the cancer data because the relationship between negative health status and hair redness was strongest in that category. We performed a logistic regression with incidence of any cancer as the dependent variable (0 = no cancer reported, 1 = cancer or precancerous lesion reported) and with three independent variables: gender, age, hair redness, and gender*hair redness interaction. For women only, incidence of any cancer was significantly associated with hair redness (OR range = 3.99, p<0.0001) and age (OR range = 12.1, p<0.0001). For men only, it was significantly associated with age (OR range = 50.7, p<0.0001) but not with hair redness (OR range = 1.63, p = 0.457). For men and women together, it was significantly associated with age (OR range = 17.7, p<0.0001) and hair redness (OR range = 3.74, p<0.0001) but not with gender (OR range = 0.67, p = 0.331) or gender*hair redness interaction (OR range = 0.47, p = 0.381). The results are shown in
The numbers above the columns show the numbers of respondents in each category.
The gender difference was greatest at the next-to-last gradation of hair redness. To learn more about this interaction between gender and gradation of hair redness, we plotted the reported mean seriousness of cancer (where 1 = no cancer reported and 6 = very serious problem with cancer) as a function of hair redness. This new variable may provide a clearer picture because it contains more information than simply the presence or absence of cancer. The results are shown in
Red hair seems to be costly for women’s health. In this study, red-haired women did worse than other women in ten health categories and better in only three. In general, women incurred more costs and gained fewer benefits from red hair than from any other hair or eye color. Brown eyes held second place but were associated with smaller negative or positive divergences in health status than those associated with red hair. Red-haired men showed a balanced pattern, doing better than other men in three health categories and worse in three. Number of children was the only category where both male and female redheads did better than non-redheads. In terms of reproductive and, ultimately, evolutionary success, red hair seems to be a plus rather than a minus.
The cancer rate was higher among red-haired women than among other women, and we initially suspected a higher rate of skin cancer as the cause. A closer look at the data, however, showed that the higher cancer rate was due not to a higher incidence of skin cancer but rather to a higher incidence of cancers in the colorectal region, the cervix, the uterus, and the ovaries (
As for the other divergences in health status associated with red hair, they too are not easily attributable to fairness of skin, and hence to UV vulnerability, again because they were reported mainly by female respondents. Although women are fairer-skinned than men, this sexual dimorphism is relatively small in fair-skinned humans and in redheads in particular, among whom both sexes are pushed up against the physiological ‘ceiling’ of skin reflectance [
Some of our findings are consistent with previous findings in the literature. Despite having more children on average, the red-haired women of this study had a higher incidence of fertility problems, which would be consistent with the higher incidence of endometriosis reported in previous studies. They also had more neurological problems, although none of these involved Parkinson’s disease. Actually, few cases of Parkinson’s would be expected, given the relatively young age of the respondents. Red-haired women showed no obvious indications of increased pain sensitivity in this study, although in some cases they might have reported more medical problems because sensitivity to pain had made them seek medical assistance more readily.
These divergences in health status thus seem to be due to a female-specific factor that is most strongly expressed in red-haired women. The relationship between this factor and hair redness seems curvilinear, i.e., female health progressively worsens on average with redder gradations of hair, but only up to a certain point. If we take the data on seriousness of cancer, the worst health status was reported by women with the next-to-last gradation of hair redness. Those with the reddest hair were actually somewhat better off (
It seems, then, that some hair and eye colors are associated with a more divergent health status, particularly in women. Red-haired women exhibit the most divergences, including a previously unreported vulnerability to colorectal, ovarian, and cervical cancer. Not all of the divergences are for the worse. In particular, red-haired women seem to enjoy greater reproductive and mating success, as measured by number of children and number of sexual partners. It may be that they have more children because they begin having them at an earlier age, although a recent study has reported that red-haired men and women lose their virginity at a later age on average [
What causes this divergent health status in red-haired women? The causation can be framed in either biochemical or evolutionary terms. First, in terms of biochemical causation, there may be a direct effect by red hair pigments (pheomelanins) or their by-products. There may also be direct effects by other genes that show linkage with genes involved in pheomelanin production. Nonetheless, such a causation would not explain why certain medical conditions occur more often in red-haired women than in red-haired men. As argued in the Introduction, the female-specific factor may be prenatal estrogen, i.e., the same factor that promotes the expression of red hair in the female fetus and brings about the higher frequency of this hair color in women. In the womb, estrogen levels are nearer the top end of the normal range for fetal development if the fetus is a female who ends up with red hair. The risk of later health problems is therefore proportionately greater.
Second, in terms of evolutionary causation, red hair may have been the last hair color to emerge in modern humans; therefore, not enough time has passed for corrective evolution, either through new alleles that produce red hair with fewer side effects or through modifier genes that neutralize the side effects of existing red hair alleles. This situation is typical of rapid evolution over relatively short spans of time [
Before proceeding to the conclusion, we should acknowledge three limitations of the present study. First, the hair and eye color data were subjective, being self-report. Less healthy individuals might have a tendency to exaggerate the redness of their hair. Second, a rather small number of respondents had the highest intensity of hair redness. To chart health problems as a function of hair redness, particularly for its highest intensity, it will be necessary to repeat our study with a larger population or one with a higher frequency of red hair. Third, the data had a high level of noise because of the subjective nature of the questions and because criteria for self-rating varied from one individual to another. As a result, even when significant correlations were found between respondent health and different factors (gender, age, hair redness, hair darkness, eye darkness, etc.), they could not explain more than a tiny proportion of total variance in health status among the respondents. We should emphasize that this tininess at least partly reflects noise in the data and does not indicate the relative importance of these factors. To provide a point of comparison, we examined respondent data on BMI and smoking, both of which strongly affect human health. Using the data in
To conclude, our findings may shed light not only on the health risks associated with red hair but also on the evolution of this highly visible color trait and, more generally, on how the diverse European palette of hair and eye colors came into being. This evolution seems to have occurred for the most part in relatively recent times, probably no earlier than the entry of modern humans into northern Europe some 30,000 years ago and no later than the oldest DNA evidence of red hair, blond hair, and blue/green eyes (Motala, Sweden), which has been dated to some 8,000 years ago [