Why complex cognitive ability increases with absolute latitude

Intelligence 46 (2014) 291–299 Contents lists available at ScienceDirect Intelligence Why complex cognitive ability increases with absolute latitude Federico R. León a,⁎, Andrés Burga León b,1 a b Universidad San Ignacio de Loyola, Humanities Faculty, Av. La Fontana 550, La Molina, Lima 12, Peru Ministerio de Educación, Measurement of Educational Quality Unit, Calle del Comercio 193, San Borja, Lima 41, Peru a r t i c l e i n f o Article history: Received 24 April 2014 Received in revised form 19 July 2014 Accepted 22 July 2014 Available online xxxx Keywords: Intelligence Latitude Evolution Cold hypothesis UVB radiation theory a b s t r a c t Evolutionary psychologists attribute the superior IQs of light-skinned populations to genetic imprints left by millenary processes promoted by cold. But a novel theory that explains IQ gains observed across recent generations ascribes them to a latitude → UVB radiation → vitamin D3 → parents' sexual hormones → family size → child's intellectual environment → IQ chain of effects. Analyses of 506,347 Peruvian children's math and reading scores from a national census confirmed that complex cognitive ability increases with absolute latitude even under tropical megathermal climates and decreases with high altitude above sea level, birth rate and social development mediate most of the effects, and reading is more strongly influenced than math. The findings weaken the evolutionary cold hypothesis and strengthen the view that contraception has the potential to reduce latitudinal IQ gaps. © 2014 Elsevier Inc. All rights reserved. 1. Introduction Light-skinned populations attain higher IQ scores than dark-skinned ones (Lynn & Vanhanen, 2002; Rushton & Jensen, 2005; Templer & Arikawa, 2006) and, since they prevail, respectively, in temperate and tropical regions of the world, absolute latitude (AL) accounts for a large part of the cognitive differences observed between nations (Lynn & Vanhanen, 2012). According to the evolutionary cold hypothesis, the predictable harsh winters of temperate regions exerted cognitive demands for survival on humans that resulted in the millenary evolvement of higher levels of intelligence; people at the tropics did not face such challenges or opportunities (Lynn, 1991; Nyborg, 2013; Rushton, 1995). Despite the theoretical, methodological, and political criticism addressed to this racial view (Asendorpf, 2007; Hunt & Sternberg, 2006; Nisbett, 2005; Volken, 2003; Wicherts, Borsboom, & Dolan, 2010; Wicherts, Dolan, & van der Maas, 2010), no competing framework emerged in the field of human ⁎ Corresponding author. Tel.: +51 1 3481262. E-mail addresses: federico.leon@usil.pe (F.R. León), andresburgaleon@gmail.com (A. Burga León). 1 Tel.: +51 1 992116587. http://dx.doi.org/10.1016/j.intell.2014.07.011 0160-2896/© 2014 Elsevier Inc. All rights reserved. intelligence but until very recently: UVB-radiation theory of latitude's influence on intelligence — URTLII (León, 2012) — attributes the AL–IQ correlations (rs) to the decreasing efficacy of skin and retina at fabricating vitamin D3 with distance from the equator, dependent, in turn, on the decaying availability of UVB photons from the equatorial line to the poles (Engelsen, Brustad, Aksnes, & Lund, 2005). Since vitamin D3 activates genes which promote production of estrogen and testosterone (Jones, Strungnell, & DeLuca, 1998; Kinuta et al., 2000), populations at high ALs present lower levels of such sexual hormones in winter (Van Anders, Hampson, & Watson, 2006; Wehr, Pitz, Boehm, März, & Obermayer-Pietsch, 2009), a cause of the seasonality of human births in temperate regions (Cummings, 2007). URTLII explains the lower rate of adolescent pregnancy in the northern than in the southern United States (Finer & Kost, 2011) and smaller total fertility rates of temperate than tropical countries (Bongaarts, 2008) as consequences of the decreasing availability of UVB photons/vitamin D3 with proximity to the poles. The resulting smaller families at the coincidentally colder habitats would produce more intelligent children because the cognitive development of the child depends on his/her intellectual stimulation at home. Since this is determined by the average mental age of parents and siblings, it decays with birth order, that is, with family size 292 F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 (Zajonc & Mullally, 1997); moreover, a child's share in the family's educational resources diminishes with birth order (Booth & Kee, 2009), thus reducing his/her opportunities for acquiring the cognitive improvements brought by formal education (Deary, 2012; Nisbett et al., 2012). Extant betweencountry IQ–birth rate rs, ranging from −0.71 to −0.85 (Lynn & Vanhanen, 2012), are consistent with these postulates. Thus, URTLII explains the Flynn effect, which refers to huge IQ gains seen from one generation to another in more than 30 nations and is interpreted as a result of the challenge of modern technologies (Flynn, 1987) or the impact of better nutrition (Lynn, 1989). In URTLII's perspective, the Flynn effect is due to the increasing use of contraception across societies and consequent reductions in family size (Bongaarts, 2008). The literature also shows that the rhythm of IQ gains is slowing down in nations where modernization began in the 19th century while it accelerates in nations where modernization began during the early to mid-20th century (Nisbett et al., 2012). URTLII accounts for this evidence considering that developed countries started their fertility transition toward smaller families earlier and are ending the transition now (Bongaarts, 2008). Evidence that AL affects income, education and women's domestic power, preference for smaller families, knowledge of family planning and use of contraceptive methods (León, 1984, 1986, 2011, 2012)2 suggests that not only family size but social development as well mediates the AL–CCA relationship. URTLII predicts additional cognitive effects of AL considering the bodily flow of dopamine, which depends on vitamin D3 (Cass, Smith, & Peters, 2006; Kesby, Eyles, Bume, & McGrath, 2011): a diminished disposition to studying, with consequences for children's cognitive development, can be assumed to occur with proximity to the equator given the greater attractiveness of playing under the elevated positive mood induced by this neurotransmitter (Siviy, 1998). The theory also predicts negative cognitive effects of altitude above sea level mediated through birth rate, for altitude increases exposure to UVB photons (Engelsen et al., 2005), thus enhancing production of testosterone (Gonzales, Gasco, Tapia, & Gonzales-Castañeda, 2009). And, since the acquisition of language occurs in the context of early parent–child relationships while math learning is more a matter of schooling, stronger effects are expected on verbal than quantitative skills. Unlike the cold hypothesis, URTLII predicts positive AL–CCA rs in any racial or climatological circumstance. The novel theory has been tested on the basis of student assessments. While scores from IQ tests and student assessments may not be exactly isomorphic within countries, rs from .77 to .94 prevail among them (Kaufman, Reynolds, Liu, Kaufman, & McGrew, 2012) and the respective latent traits correlate above .80 (Sonnleitner, Keller, Martin, & Brunner, 2013); this justifies encompassing both into the larger concept of complex cognitive ability (CCA), which has international validity (Rindermann, 2007). León (in press) utilized two data sources: (a) Peru Ministry of Education's reported mean scores per region of education entailing the academic achievements of children in second year of primary instruction (N = 24 regions, math and reading scores averaged) and (b) mean PISA (Programme for International Student Assessment) scores in science among 15 year-old students in various American countries (N = 11 countries). Despite the tiny sample sizes, AL emerged as a consistent determinant of student scores in both studies. According to the Baron and Kenny (1986) mediation analysis model applied to the Peruvian data, AL's influence on student assessment scores was mediated through total fertility rate. Ordinary least squares regression for countries revealed that racial composition disappeared as a determining factor when AL was controlled whereas the influence of latitude remained significant when race was controlled. This article presents the results of a more rigorous and complex test. 2. Method 2.1. Context Peru is a country in the southern hemisphere extending from 0° 02′ S to 18° 21′ S where 45% of the population are Amerindians whose ancestors arrived from the Northeast Asia more than 14,000 years ago and whose skin is lighter than the African skin and darker than the European skin. Other subpopulations are European-descended (15%) and descendants of African slaves or Chinese or Japanese immigrants (3%); Amerindians, Europeans, and Africans have contributed importantly to the mestizo gene pool (37%) (Putterman & Weil, 2011). Lighter skin is not more prevalent alongside distance from the equator in Peru. Rather, the linguistics literature suggests the opposite: whereas Spanish is the mother tongue of eight out of every 10 Peruvians, Amerindian languages have survived to a greater extent in the south than in the center of the country and are virtually extinct in the north except for circumscribed areas and small and scattered Amazonian tribes (Knapp, 1987). 2.2. Subjects The data originated in a national census carried out in 2011 by the Peru Ministry of Education that obtained math and reading scores from 506,347 children in 2nd grade of primary instruction, i.e., aged about eight (51% boys, 49% girls). The Ministry yearly evaluates student performance at the national level since 2007; the 2011 census was conducted on November 29 and 30. The Ministry assigns schools of a national listing to specially trained teams of data collectors who travel to the respective sites and give the testing materials to students assembled in groups in classrooms. The Ministry targeted 1769 of the 1837 Peruvian districts (where there were schools with more than five students); Fig. S12 and Table S12 present relevant information. Of the targeted districts, 1479 satisfied the census' coverage standards (at least 90% of schools accessed and 80% of students tested), 259 did not, and 31 were not reached. 2.3. Measurements 2.3.1. Complex cognitive ability We submitted the student scores to psychometric analyses. Only two items of the math test and two items of the reading test did not adjust to Rasch models and failed in tests of unidimensionality (see Tables S2–S4).2 The reliabilities of the 2 See Online Supplement. F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 293 math and reading total scores, respectively .89 and .87, indicate a high degree of consistency and precision in such scores. The math and reading scales correlated .89. 2.3.2. Physical variables Peru's Ministry of Education website provides geographic locators for each Peruvian town containing a school, including longitude, latitude, and altitude. We identified the town with the greatest student population within a district and assigned its latitude and altitude to the district. In a dozen cases, latitude and/or altitude information were unavailable at the Ministry's website and we had to calculate it using maps. Access to sea was measured using the map of Fig. S12 and assigning 3 points to administrative regions of education at the Pacific Ocean littoral (310 districts), 2 points to regions separated from the sea by one region (728 districts), and 1 point to those separated from the sea by two or more regions (700 districts). 2.3.3. Social variables The student census database supplied information on the number of children enrolled at a school, whether the school was private or public, whether it had one or several teachers, and the proportion of female students. Income, education, and life expectancy were reported at district level by the Peru 2009 Human Development Report (PNUD, 2010). Per capita family income of the report was calculated in national currency (actually, expenses). Educational achievement per district of the report is an average of literacy and school attendance; the latter was calculated as the percentage of the population between five and 18 years of age attending school. Life expectancy of the report was calculated considering the infant mortality rate. We obtained from the 2007 National Census (INEI, 2008) a family size index calculated per district as the average number of children ever had by men and women. To address social development, we extracted a principal component from eight correlated variables (see Table S5)2 – district's distance from sea, income, education, life expectancy and its schools' number of students, proportion private, proportion with only one teacher, and proportion of female students (see Table S6)2 – and calculated social development scores using regression analysis. Income was by far the main contributor to this score. 2.4. Regions Peru's tropical condition is moderated by cold brought by the marine Humboldt Current onto its Pacific coast from Antarctica and by the altitude of the Andes mountains that cross the country from Bolivia to Ecuador (Fig. S2A).2 The country presents 26 Thornthwaite climates (Fig. S2C).2 The Peruvian Amazonía (black and three large gray areas in eastern part of Fig. 1) is a well-defined ecological region (see Fig. S2B2) that contains the four humid/megathermal climates of Peru making up the Peruvian rainforest (Brack, 1993). It has a maximal 600 meters above sea level and a 26 °C average yearly temperature. Most of it is a very rainy region, the remainder only rainy; one sub-region presents greater atmospheric humidity than the others and one presents rain deficiency in winter. Due to the Amazonía's difficulties for agriculture, it is underpopulated, overurbanized, and poorly connected to the remainder of the country (Fig. S3).2 The coastal strip defined for Fig. 1. Regions defined for the research (see text). this research (green in Fig. 1; see Fig. S22 about its construction) presents a homogeneous mild climate and a maximum altitude of 1000 m and is crossed by 40 small rivers forming rich tiny valleys amid the Pacific desert. It is the more industrialized and urbanized region of Peru and contains the largest cities and the richest and most educated sub-population of the country, although its cultural homogeneity is disrupted by intensive migration from other regions. The remainder of the country (colored space between the coastal strip and the Amazonía in Fig. 1) is a rather heterogeneous territory. Its schools are located from 4 to 5066 m of altitude and its climate ranges from arctic cold to arid megathermal. This region contains most of the Amerindian rural population of Peru. Its higher altitude is associated with lower CCA scores and the effects of altitude appear to be mediated through social variables (León & Avilés, 2013). 2.5. Analytic strategy The units of analysis for our main analyses were the 1479 districts which met the census' coverage standards (90% of listed schools accessed and 80% of registered students tested). Data from the 259 districts which failed to meet the standards were then incorporated and submitted to supplementary 294 F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 analyses. The analyses utilized employed bootstrapping (1000 samples). 3. Results 3.1. Findings at the national level The mean CCA scores found among girls and boys at the national level (Fig. 2) were consistent with the study hypotheses and with the literature that shows male advantages in math and female advantages in verbal abilities (Guiso, Monte, Sapienza, & Zingales, 2008). The scores showed a significant dependence on AL under various regression models, except for math when social development was controlled; and AL influenced reading more strongly than math (Table 1). It can be noticed that, whereas the AL–CCA regression coefficients increased when altitude was set constant, they decreased when the variables controlled were birth rate and social development. We treated more explicitly AL and altitude as exogenous variables and the other two as endogenous ones in path analyses whose results are presented in Fig. 3. Since the previous analyses had indicated equivalence of the boys' and girls' scores regarding each variable, gender was ignored. The results of the path analyses were consistent with the study hypotheses on birthrate's mediation of the AL–CCA relationship and altitude's effects on birth rate, although rather moderately; direct cognitive effects of AL predicted from the assumed flow of dopamine were observed regarding reading. Altitude's strong negative influence on social development reflects the difficulties for agriculture and the construction of buildings, roads and Table 1 Standardized coefficients from the regression of math and reading scores on four predictors in Peru, per gender, type of ability, and regression model. Predictors Math girls Absolute latitude Altitude Birth rate Social development Math boys Absolute latitude Altitude Birth rate Social development Reading girls Absolute latitude Altitude Birth rate Social development Reading boys Absolute latitude Altitude Birth rate Social development Model I Model II Model III Model IV .064⁎ .182⁎⁎⁎ −.280⁎⁎⁎ .127⁎⁎⁎ −.184⁎⁎⁎ −.244⁎⁎⁎ .026 .039 −.134⁎⁎⁎ .400⁎⁎⁎ .065⁎ .184⁎⁎⁎ −.283⁎⁎⁎ .135⁎⁎⁎ −.195⁎⁎⁎ −.224⁎⁎⁎ .027 .044 −.106⁎⁎⁎ .428⁎⁎⁎ .103⁎⁎⁎ .290⁎⁎⁎ −.442⁎⁎⁎ .223⁎⁎⁎ −.327⁎⁎⁎ −.297⁎⁎⁎ .077⁎⁎⁎ −.005 −.138⁎⁎⁎ .577⁎⁎⁎ .114⁎⁎⁎ .283⁎⁎⁎ −.401⁎⁎⁎ .216⁎⁎⁎ −.282⁎⁎⁎ −.304⁎⁎⁎ .069⁎⁎ 044. −.142⁎⁎⁎ .584⁎⁎⁎ Note. Ns = 1471 to 1479 districts which met the census coverage standards (90% of listed schools accessed and 80% of registered students tested). ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001, two-tailed. bridges posited by rugged terrain as well as the fact that Amerindians at high Peruvian altitudes still maintain agricultural practices of the 16th century. Social development's mediation of AL–CCA effects through its influence on birth rate is consistent with prior Peruvian evidence (León, 1984, 1986, 2011).2 But social development mediated the influence of AL on CCA independently of birth rate, too, which can be understood considering that poverty affects cognitive functioning through negative emotional states (Mani, Mullainathan, Shafir, & Zhao, 2013) and that soil erosion and tropical diseases increasingly impoverish populations with proximity to the equatorial line (Bloom & Sachs, 1998). The use of social development factor scores – versus using income, its strongest determinant, alone – is justified by previous findings in Peru2 and by the fact that income and factor scores from the other components of social development made independent contributions to birth rate in this study (Table S8).2 3.2. Findings at subnational level Fig. 2. Mean complex cognitive ability score, per type of ability, gender, and latitude level in Peru. Districts were the units of analysis. We divided the country into three equal latitudinal segments. Observed limits of the segments were: North (N), 0.37°–5.39° S (N = 155 districts). Center (C), 5.90°–11.80° S (N = 745 districts). South (S), 11.80°–18.06° S (N = 838 districts). Data come from districts that fully achieved the census' coverage standards. The robustness of the findings and the temperature issue were addressed in the more homogeneous settings of the Peruvian Amazonía and the coastal region, as well as in the remainder of Peru. In contrast with the national level, the CCA increases with AL were not monotonic; however, all the linear trends were positive, except for math in the remainder of Peru (Fig. 4). AL influenced CCA through birth rate in the Amazonía and its effects there were stronger on reading than math (Fig. 5A). Birth rate's effects on CCA were greater than those seen in the other Peruvian regions (Fig. 5A, B), which suggests that the negative cognitive effects of family size are intensified under strong UVB radiation. The Amazonía is the Peruvian region closest to the equator and, given its high temperature, its F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 295 Fig. 3. Standardized path coefficients for Peru from saturated model. Districts were the units of analysis. Only those which met the census' coverage standards were considered (N = 1471). The r between AL and altitude owes to the increasingly wider and higher Andes mountains with absolute latitude in Peru. + p b .10, *p b .05, **p b .01, ***p b .001. residents wear slight clothing and expose large parts of the body to sunlight all year round. Regional stereotyping of Amazonians in Peru, pointing to sexual hyperactivity and promiscuity,3 suggests deficits of parenting, which could help explain an incongruous finding of the study: whereas the Amazonía occupied an intermediate position between the rich coast and poor remainder of the country in social development, it fell behind the latter in CCA (Table S7).2 Consistent with the emerging explanatory hypothesis on parenting, the effects of birth rate on CCA decayed from northern to central to southern latitudinal segments when we divided the Amazonía into three equal parts (Figs. S4, S5).2 Similar was the case in the coast (Figs. S6, S7)2 and remainder of Peru (Figs. S8, S9).2 Altitude's positive direct influence on CCA probably reflects children's greater exposure to disease – hence, school absenteeism – at lower Amazonía altitude. The observed negative effect of altitude on birth rate invites speculation on the relationship between production of cocaine at 400 m. and above in the Amazonía (United Nations & Devida, 2005) and the fact that the use of this drug impairs fertility (Fronczak, Kim, & Barqawi, 2012). The strongest effects of AL on birth rate were observed in the Peruvian coast (Fig. 5B), which is exposed to weaker UVB radiation than the Amazonía and remainder of the country due to the filtering of UVB photons most of the year by the clouds and fog caused by the cold Humboldt current and the pollution 3 These views are consistent with the high levels of sexually transmitted infection and HIV risk behaviors observed in large Amazonian cities as well as in remote villages (Cáceres, Mendoza, Yon, Rosasco, & Cabezudo, 1998; National Institute of Statistics and Informatics, 2001; Paris et al., 2001; Zavaleta et al., 2009). generated by its industry and transport; moreover, the coast's high level of urbanization induces more time spent under roof and its cold in winter makes people wear heavy clothing in wintertime (León, 2012). Hence, it should not be surprising that coastal children achieved the highest CCA scores of the country (Table S7),2 albeit the coast's social development certainly helped. The stronger effect of AL on birth rate seen in this region may imply that the differential impact of UVB radiation on sexual life is greater when lower general levels of vitamin D3 are involved. This probably is related to another coastal characteristic: the emergence of direct effects of AL on CCA, which did not occur in the Amazonía. This finding suggests that greater differences take place between northern and southern children regarding proclivity to playing versus studying under the lower levels of flowing dopamine expected in the Peruvian coast. The implication of this explanatory hypothesis – that the direct influence of AL on CCA should be greater at the southern than at the northern Peruvian coast – was upheld in a rough comparison between latitudinal segments (Fig. S7).2 Altitude was irrelevant to birth rate in the coastal region and exhibited opposite effects in the Amazonía and the remainder of Peru. The positive influence observed in the latter region may indicate that, as a determinant of sexual activity, greater exposure to UVB photons is effective past the 1000 m. The direct negative effect of altitude on CCA seen in the remainder of Peru probably shows that 14,000 years of adaptation have not been enough and Andean children's cognitive functioning is still affected by the hypoxia of very high altitude (Beall, 2006). This contradicts the León and Avilés (2013) findings of full mediation of the altitude–CCA relationship through social development; but their study was limited to the 8° S–10° S latitudinal segment and our analytic techniques were stronger. 296 F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 Fig. 4. Mean complex cognitive ability score, per type of ability and latitude level. (A) Amazonía, N = 119. (B) Coastal region, N = 253. (C) Remainder of Peru, N = 1107. We divided each region into six equal latitude segments. The linear trends are depicted. 3.3. Constraints The strength of the AL–CCA relationships observed in Peru as a whole (rs = .08 for math and .14 for reading) was considerably weaker than that observed in the international studies (rs ranging from .68 to .72; Lynn & Vanhanen, 2012); the significance of the findings owes to the high number of cases (highest N = 1479). This can be attributed in part to the complexity of the Peruvian territory: the AL-averaged reading– math r was considerably higher for the homogeneous research regions (.42 in the homogeneous Amazonía and .27 in the homogeneous coast versus .05 in the heterogeneous remainder of the country). On the other hand, the latitudinal span covered in the international studies was 63° from Uganda through Finland vis-à-vis only 17.7° encompassed in Peru from the northernmost observation to southernmost observation. The findings reported in Fig. 5 emerged along the effective 12.88° of the Amazonía, 14.38° of the coast, and 14.42° of the remainder of the country. They remained robust when we combined the north and center segments of the Amazonía, but AL did not influence birth rate at the center-south segment combination (Fig. S4).2 The results were robust at the north-center and center-south segments of the coast (Fig. S6)2 and remainder of the country (Fig. S8)2 as well as at the north, center, and south segments of the remainder of the country (Fig. S9),2 but not at the center segment of the coast (Fig. S7)2 neither the north nor the center segments of the Amazonía (Fig. S5).2 The geography and population of a short latitudinal segment might have to be extremely homogeneous to yield significant AL–CCA or AL– birth rate–CCA path coefficients. In contrast, robust effects of social development on birth rate and CCA were seen at each latitudinal segment (Figs. S4–S9);2 this is so because social development depends on sources of variance which are effective regardless of AL. It should also be remarked that the evidence reported came from the 1479 Peruvian districts which achieved the census' coverage standards (Table S1).2 This focus responded to our desire of making sure that the registered latitude of the district accurately reflected the traits of its population. Important loss of schools and students at 259 districts and the loss of 31 districts are attributable to geographic barriers to access and the fact that the days of the week when some remote rural schools receive visiting teachers did not coincide with the moment of the 2-day census' field activities. The districts with deficient coverage were more frequent in the Amazonía (25%) and the remainder of the country (16%) and negligible in the coastal region (3%); therefore, our findings based on coastal data may be stronger than those based on data from the other regions. On the other hand, missing schools and students were more frequent in northern than southern Peru and the districts with deficient coverage exhibited poorer CCA scores (Table S92), although not enough to modify the differences between regions (Table S10).2 Consequently, the findings resulted slightly more favorable to the theory when the 259 districts which did not achieve the coverage standards were included in the analyses (compare Tables S5 with S11, Tables 1 with S12 and Figs. 3 and 5 with S10 and S11);2 that is, the missing cases ran against the odds of spuriously upholding URTLII's hypotheses. 4. Discussion Rather weak effect sizes of AL on birth rate and of the latter on CCA prevailed in the path analyses of this study, especially compared to the strength of the social development–CCA relationships. In our view, however, this is less important than the fact that the Peruvian findings were consistent with those generated by evolutionary psychologists throughout the world F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 Fig. 5. Standardized path coefficients for the Peruvian Amazonía (A), coastal region (B), and remainder of the country (C) from saturated models. (Lynn & Vanhanen, 2012) despite the short latitudinal segments targeted. The evolutionary evidence has been downplayed on methodological grounds (Hunt & Sternberg, 2006; Nisbett, 2005; Volken, 2003; Wicherts, Borsboom, & Dolan, 2010; Wicherts, Dolan, & van der Maas, 2010), but, since our measurements were reliable and the data represented the targeted population reasonably well, the study findings are free of the methodological criticism raised against past evidence. Hence, they strengthen URTLII's postulate of universality of 297 the positive relationship existing between AL and CCA, i.e., between- and within-races and climates. They also render credible the AL–CCA relationship's contemporary mediation through family size and social development. Do they uphold the rationales underlying our predictions concerning AL's direct effects on CCA, differential effects on reading versus math, and negative cognitive effects of altitude? Fig. S122 depicts the URTLI links that were confirmed in our study and the theoretical links that have been reported in the literature. Evidently, URTILI needs further empirical strengthening through direct measurements of UVB radiation, vitamin D3, and sexual hormones as well as demonstration of the relationships between these variables and AL and IQ within saturated path models. We used AL as a proxy for the real thing because available satellite measurements of UVB radiation require adjustments for altitude, clouds, pollution, soil reflectance and other variables and this information is provided by land stations which are scarce in Peru (N = 9) and do not provide coverage even for the whole of Metropolitan Lima. Nonetheless, URTLII's validity is likely to be directly testable in the near future; an increased building of land stations can be expected worldwide given national governments' concern with climate change. McKenzie, Ben Liley, and Björn (2009) have advocated intensive measurement of UVB radiation's effects on vitamin D3 levels of populations as a complement to the present focus on the amount of UVB radiation needed to cause erythema. URTLII also has implications for the climate change topic; it predicts a worldwide upsurge of sexual activity and reduction of depressive moods in addition to the anticipated increase in the rate of skin cancer. The findings weaken the evolutionary cold hypothesis. Since skin color is not darker nearer the equator in Peru (Knapp, 1987) and cold is and was absent in low-land tropics even during the Last Glacial Maximum (Gasse, Chalié, Vincens, Williams, & Williamson, 2008), the evolutionary framework predicts zero AL–CCA rs among Peruvian children or under megathermal climates; this was contradicted by the positive relationships found in the study. The observed mediation of the AL–CCA rs through birth rate is not explainable by cold, either, despite evolutionary psychologists' insights entailing sexual activity along AL: in one of their theoretical perspectives, adaptations to cold would have included slower life histories associated with a reproductive strategy of restrained sexuality (Rushton, 1995) and, in a second evolutionary framework, higher IQ–lower testosterone coevolution would have taken place as an adaptation to biological demands for allocating energy under extreme cold (Nyborg, 2013). What these evolutionary theories of cold postulate is the inheritance of increasing levels of IQ and sexual restraint or IQ and testosterone underproduction alongside AL, not that sexual restraint or testosterone perform as mediators of the AL–CCA relationship. A third evolutionary theory states that what we now call general intelligence originally evolved as a domainspecific adaptation to solve the evolutionarily novel problems that Homo sapiens found upon leaving the African savanna, cold being just one of them; other evolutionary novel problems would have included altitude above sea level (Kanazawa, 2008). However, our findings at 4 to 5066 m of altitude revealed negative rather than positive cognitive effects of this variable. In turn, the dearth of evidence on the heritability of verbal versus quantitative skills impedes interpreting in an evolutionary-cold perspective the differences found between 298 F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 math and reading regarding susceptibility to AL's influence. Finally, temperature cannot account for the key relationship between AL and skin color, which UVB radiation does: the skin of modern humans evolved from dark to light as an adaptation to the scarcity of UVB photons encountered in Europe and Northern Asia about 45,000 years ago; less melanin (i.e., pigment), which refracts sunlight, facilitated conversion of weak UVB radiation into vitamin D3 (Jablonski & Chaplin, 2010). This, however, was not enough, and thus Caucasians still experience a decline in vitamin D3 alongside latitude (Hagenau et al., 2009). The weakening of the evolutionary-cold hypothesis does not necessarily imply that evolutionary concepts should be discarded in the explanation of the AL–CCA relationship. In fact, URTLII may need an evolutionary dimension to be able to more fully explain differences in intelligence between populations over millennia (Lynn, 2009). A more specific research need in a historic perspective entails the role of latitude in the formation of social values associated with investments in parenting and intellectual effort and how these values interact with the new climates faced by migrant populations, such as Europeans in South America and Chinese in Singapore. The possible moderation of the AL–CCA relationship by the extent of pigment in the skin should be assessed, too; greater production of testosterone appears to be associated with light- than dark-skin in the tropics (Beall et al., 1992) and similar should be the case of estrogen. Moreover, since heavily pigmented African American populations are prone to vitamin D3 deficit (Signorello et al., 2010), they can be expected to present smaller AL–CCA rs than Caucasians in temperate North America. Our closing point puts the skin-color–IQ relationship into a dynamic perspective. Since increased contraceptive use is likely to reduce family size to a greater extent among the more sexually active populations and less intelligent populations enroll later in family planning, contraceptive use functions as an equalizer of family size as more families enroll in contraception. Thus, our findings strengthen the view that increased contraceptive use resulting from spontaneous social processes or promoted by family planning agencies can be expected to diminish current skin-color- and latitudinal-IQ gaps between- and within-countries (León, 2012, in press). Acknowledgments Funds for this research were provided by the Research Center at Universidad San Ignacio de Loyola, Lima, Peru. We are indebted to Liliana Miranda, Director of the Peru Ministry of Education's Measurement of Educational Quality Unit, for lending us the Peru census database; Edvar Avilés for his help in data entering and analysis, Gustavo Gonzales and Manolete Moscoso for their comments on an earlier version of the paper, Nina Bustamante for her support, and an anonymous reviewer for his/her suggestions. FRL designed the study, drafted the manuscript, performed part of the analyses, and interpreted the findings. ABL performed part of the analyses and contributed to the design of the study and interpretation of the findings. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.intell.2014.07.011. References Asendorpf, J. B. (2007). The Big G-factor of national cognitive-ability comparisons: Not trivial and not immutable. European Journal of Personality, 21, 709–712. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Beall, C. (2006). Andean, Tibetan, and Ethiopian patterns of adaptation to highaltitude hypoxia. Integrative and Comparative Biology, 46, 1,18–1,24. Beall, C. M., Worthman, C. M., Stallings, J., Stroll, K. P., Brittenham, M., & Barragan, M. (1992). Salivary testosterone concentration of Aymara men native to 3600 m. Annals of Human Biology, 19, 67–78. Bloom, D. E., & Sachs, J.D. (1998). Geography, demography, and economic growth in Africa. Brookings Papers on Economic Activity, 415, 207–295. Bongaarts, J. (2008). Fertility transitions in developing countries: Progress or stagnation. Studies in Family Planning, 39, 105–110. Booth, A. L., & Kee, H. J. (2009). Birth order matters: The effect of family size and birth order on educational attainment. Journal of Population Economics, 22, 367–397. Brack, A. (1993). El país en el que vivimos. Lima: Editorial Salesiana. Cáceres, C., Mendoza, W., Yon, C., Rosasco, A., & Cabezudo, C. (1998). SIDA en el Perú: Imágenes de diversidad, situación y perspectivas de la epidemia en Chiclayo, Cuzco e Iquitos. Lima: UPCH y REDES Jóvenes. Cass, W. A., Smith, M. P., & Peters, L. E. (2006). Calcitriol protects against the dopamine- and serotonin-depleting effects of neurotoxic doses of methanphetamine. Annals of the New York Academy of Sciences, 1074, 261–271. Cummings, D. R. (2007). Additional confirmation for the effects of environmental light intensity on the seasonality of human conceptions. Journal of Biosocial Science, 39, 383–396. Deary, I. J. (2012). Intelligence. Annual Review of Psychology, 63, 453–482. Engelsen, O., Brustad, M., Aksnes, L., & Lund, E. (2005). Daily duration of Vitamin D synthesis in human skin with relation to latitude, total ozone, altitude, ground cover, aerosols and cloud thickness. Photochemistry and Photobiology, 81, 1287–1290. Finer, L. B., & Kost, K. (2011). Unintended pregnancy rates at the state level. Perspectives on Sexual and Reproductive Health, 43, 78–87. Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101, 171–191. Fronczak, C. M., Kim, E. D., & Barqawi, A. B. (2012). The insults of illicit drug use on male fertility. Journal of Andrology, 33, 515–528. Gasse, F., Chalié, A., Vincens, A., Williams, M.A. J., & Williamson, D. (2008). Climatic patterns in equatorial and southern Africa from 30,000 to 10,000 years ago reconstructed from terrestrial and near-shore proxy data. Quaternary Science Reviews, 27, 2316–2340. Gonzales, G. F., Gasco, M., Tapia, V., & Gonzales-Castañeda, C. (2009). High serum testosterone levels are associated with excessive erythrocytosis of chronic mountain sickness in men. American Journal of Physiology — Endocrinology and Metabolism, 296, 1319–1325. Guiso, L., Monte, F., Sapienza, P., & Zingales, L. (2008). Culture, gender, and math. Science, 320, 1164–1165. Hagenau, T., Vest, R., Gisell, T. N., Poulsen, C. S., Erlandsen, M., Mossekilde, L., et al. (2009). Global vitamin D levels in relation to age, gender, skin pigmentation and latitude: An ecologic meta-regression analysis. Osteoporosis International, 20, 113–140. Hunt, E., & Sternberg, R. J. (2006). Sorry, wrong numbers: An analysis of a study of a correlation between skin color and IQ. Intelligence, 34, 131–137. INEI (2008). Censos Nacionales 2007: XI de Población y VI de Vivienda. Lima: Instituto Nacional de Estadística e Informática. Jablonski, N. J., & Chaplin, G. (2010). Human skin pigmentation as an adaptation to UV radiation. Proceedings of the National Academy of Sciences of the United States of America, 107, 8962–8968. Jones, G., Strungnell, S. A., & DeLuca, H. F. (1998). Current understanding of the molecular actions of vitamin D. Physiology Review, 78, 1193–1231. Kanazawa, S. (2008). Temperature and evolutionary novelty as forces behind the evolution of general intelligence. Intelligence, 36, 99–108. Kaufman, S. B., Reynolds, M. R., Liu, X., Kaufman, A. S., & McGrew, K. S. (2012). Are cognitive g and academic achievement g one and the same g? An exploration on the Woodcock–Johnson and Kaufman tests. Intelligence, 40, 123–138. Kesby, J. P., Eyles, D. W., Bume, T. H. J., & McGrath, J. J. (2011). The effects of vitamin D on brain development and adult brain function. Mollecular and Cellular Endocrinology, 347, 121–127. Kinuta, K., Tanaka, H., Morikawe, T., Aya, K., Kato, S., & Seino, Y. (2000). Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology, 141, 1317–1324. Knapp, G. (1987). Linguistic and cultural geography of contemporary Peru. University of Texas at Austin Institute for Latin American Studies Papers, 1–19. F.R. León, A. Burga León / Intelligence 46 (2014) 291–299 León, F. R. (1984). El eje fecundatorio norte-sur del Perú: Una interpretación psicológica. Revista de Psicología, 2, 95–111. León, F. R. (1986). Factores psicosociales, psicoeconómicos, y psicosexuales en el eje fecundatorio norte-sur del Perú. In F. R. León (Ed.), Psicología y realidad peruana: El aporte objetivo (pp. 87–105). Lima: Mosca Azul Editores. León, F. R. (2011). Latitud sur y control económico del hogar por la mujer peruana. Revista de Psicología, 29, 362–388. León, F. R. (2012). The latitudinal tilts of wealth and education in Peru: Testing them, explaining them, and reflecting on them. Economia, 35, 60–102. León, F. R. (2014s). Efectos de la latitud en el logro escolar: ¿Evolucionarios o vía la radiación ultravioleta contemporánea? In R. León, J. Ramírez, & D. Jáuregui (Eds.), Lima: Universidad Ricardo Palma (in press). León, F. R., & Avilés, E. (2013). Efectos de la altitud sobre la habilidad cognitivas compleja. Propósitos y Representaciones, 1(2), 31–56. Lynn, R. (1989). Positive correlation between height, head size and IQ: A nutrition theory of the secular increases in intelligence. British Journal of Psychology, 59, 372–377. Lynn, R. (1991). The evolution of race differences in intelligence. Mankind Quarterly, 32, 99–173. Lynn, R. (2009). Consistency in race differences in intelligence over millennia: A comment on Wicherts, Borsboom and Dolan. Personality and Individual Differences, 48, 100–101. Lynn, R., & Vanhanen, T. (2002). IQ and the wealth of nations. Westport, CT: Praeger. Lynn, R., & Vanhanen, T. (2012). National IQs: A review of their educational, cognitive, economic political, demographic, sociological, epidemiological, geographic and climatic correlates. Intelligence, 40, 226–234. Mani, A., Mullainathan, S., Shafir, E., & Zhao, J. (2013). Poverty impedes cognitive function. Science, 341, 976–980. McKenzie, R. L., Ben Liley, J., & Björn, L. O. (2009). UV radiation: Balancing risks and benefits. Photochemistry and Photobiology, 85, 88–98. National Institute of Statistics and Informatics (2001). Reproductive and Health Survey, Peru 2000. Lima: NISI. Nisbett, R. E. (2005). Heredity, environment, and race in IQ: A commentary on Rushton and Jensen (2005). Psychology, Public Policy, and Law, 11, 302–310. Nisbett, R. E., Aronson, J., Blair, C., Dickens, W., Flynn, J., Halpern, D. F., et al. (2012). Intelligence: New findings and theoretical developments. American Psychologist, 67, 130–159. Nyborg, H. (2013). Migratory selection for inversely related covariant T-, and IQNexus traits: Testing the IQ/T-Geo-Climatic-Origin theory by the General Trait Covariance model. Personality and Individual Differences, 55, 267–272. Paris, M., Gotuzzo, E., Goyzuetas, G., Aramburú, J., Cáceres, C., Crawford, D., et al. (2001). Motorcycle taxi drivers and sexually transmitted infecions in a Peruvian Amazon city. Sexually Transmitted Diseases, 28, 11–13. 299 Putterman, L., & Weil, D. (2011). World migration matrix, 1500–2000. Retrieved April, 25 2011 from. www.econ.brown.edu/fac/louis_putterman Rindermann, H. (2007). The g-factor of international cognitive ability comparisons: The homogeneity of results in PISA, TIMSS, PIRLS and IQ tests across nations. European Journal of Personality, 21, 667–706. Rushton, J. P. (1995). Race, evolution, and behavior: A life history perspective. New Brunswick: Transaction. Rushton, J. P., & Jensen, A.R. (2005). Thirty years of research on race differences in cognitive ability. Psychology, Public Policy, and Law, 11, 235–294. Signorello, L. B., Williams, S. M., Zheng, W., Smith, J. R., Long, J., Hargreaves, M. K., et al. (2010). Blood vitamin D levels in relation to genetic estimation of African ancestry. Cancer Epidemiology, Biomarkers & Prevention, 19, 2325. Siviy, S. M. (1998). Neurobiological substrates of play behavior: Glimpses into the structure and function of mammalian playfulness. In M. Bekoff, & J. A. Biers (Eds.), Animal play: Evolutionary, comparative, and ecological perspectives. Cambridge: Cambridge University Press. Sonnleitner, P., Keller, U., Martin, R., & Brunner, M. (2013). Students' complex problem-solving abilities: Their structure and relations to reasoning ability and educational success. Intelligence, 41, 289–305. Templer, D. L., & Arikawa, H. (2006). Temperature, skin color, per capita income, and IQ: An international perspective. Intelligence, 34, 121–139. United Nations & Devida (2005). Peru: Coca cultivation survey. Lima, Peru: Authors. Van Anders, S. M., Hampson, E., & Watson, N. V. (2006). Seasonality, waist-tohip ratio, and salivary testosterone. Psychoneuroendocrinology, 31, 895–899. Volken, T. (2003). IQ and the wealth of nations. European Sociological Review, 19, 411–412. Wehr, E., Pitz, S., Boehm, B. O., März, W., & Obermayer-Pietsch, B. (2009). Association of vitamin D status with serum androgen levels in men. Clinical Endocrinology, 73, 243–248. Wicherts, J. M., Borsboom, D., & Dolan, C. V. (2010). Why national IQs do not support evolutionary theories of intelligence. Personality and Individual Differences, 48, 91–96. Wicherts, J. M., Dolan, C. V., & van der Maas, J. L. J. (2010). The dangers of unsystematic selection methods and the representativeness of 46 samples of African test-takers. Intelligence, 38, 30–37. Zajonc, R. B., & Mullally, P. R. (1997). Birth order: Reconciling conflicting effects. American Psychologist, 52, 685–699. Zavaleta, C., Fernández, C., Konda, K., Valderrama, Y., Vermund, S. H., & Gotuzzo, E. (2009). Short report: High prevalence of HIV and syphilis in a remote native community of the Peruvian Amazon. American Journal of Tropical Medicine Hygiene, 76, 703–705.