Intelligence and the Environment

Environmental factors (positively) affecting IQ

• Cognitively stimulating environment

  • Equality across society?

• Optimal health (e.g. good nutrition and medical care)

• Avoiding adversity/trauma

• Avoiding environmental toxins/pollutants

• Societal differences (e.g. westernisation, technology)

Coutrot et al. (2022)

Sea Hero Quest - Mobile game to test spatial ability in humans

• Tested 2.5 million people from every country in the world

• Spatial ability = Correlated with economic health

• Gender inequality = Predictive of gender differences in navigation ability

Coutrot et al. (2022) - Structure of Environment

Sea Hero Quest - Wayfinding performance:

• Lowest for people living in cities

• Higher for people living in suburbs, mixed and rural areas

• Decreased with age

Coutrot et al. (2022) - Entropy

People were better at navigating in environments that were topologically similar to where they grew up.

• Growing up in cities with a low street network entropy (simple, grid-like street system e.g. Chicago)

  • Better results at video game levels with a regular layout,

• Growing up outside cities or in cities with a higher street network entropy (complex street network e.g. Prague)

  • Better results at more entropic video game levels.

Evidence for the effect of the environment on cognition on a global scale, and the importance of urban design

First-born advantage

• Parents move from parenting one child to multiple children

• Systematic shifts in parental behaviour and the home environment.

• Differences may be of interest, consistent patterns across families would be difficult to explain in terms of genetic variation

The National Longitudinal Survey of Youth 1979 (NLSY79)

A nationally representative sample of 12,686 American youths born between 1957-64.

First interviewed in 1979 (aged 14-22 years).

Periodic surveys:

• Employment

• Income

• Welfare program participation

• Education

and other background variables.

Children of the NLSY79 (CNLSY79)

In 1986, 11,420 children of the 6,283 female NLSY79 respondents = interviewed biannually

Child survey:

• Scores from cognitive and non-cognitive assessments

• Prenatal investments

• Birth outcomes

• Early childhood parental investments

• Health and quality of the home environment

As well as additional demographic and development information.

Lehmann et al. (2014) Non-cognitive Associations

CNLSY79 Sample

Later-born children:

• Similar health and development.

• Differences in maternal behaviour during pregnancy & first year of life

  • (i.e. alcohol and tobacco consumption, and decision to breastfeed).

Lehmann et al. (2014) Cognitive Associations

• As early as 1 year: later born children score lower on cognitive tests

• The gap increases until school entry, then becomes statistically irrelevant

• Explained by cognitive stimulation by mothers

  • Only evident for cognition

Angelson et al. (2001)

• Breastfeeding for less than 3 vs 6 months

• 3 months → increased risk of low test scores below the median value on scales of infant development (13m) and IQ (5y)

Mortensen et al., (2002)

Assessed with 2 different intelligence tests.

2 independent samples of young adults

• A significant positive association between the duration of breastfeeding and intelligence was observed

• Independent of a wide range of possible confounding factors

Deoni et al. (2013) - High-order Cognition

Breastfed children at age 2

Improved receptive language scores and enhanced development in key parts of the brain associated:

• Executive functioning

• Planning

• Social-emotional functioning

• Language

Breastfeeding alone > Combination of breastfeeding and formula > Formula alone.

Deoni et al. (2013) - Breastfeeding Duration and Myelin Content

• n=143

• Extended breastfeeding (>15 months) was positively associated with greater myelination in somatosensory, auditory and language areas.

Victora et al. (2015) - Method

Long-term Consequences of Breastfeeding on IQ

Years of schooling and income at age 30 yrs.

Information from 3493 individuals on IQ and breastfeeding (in Brazil).

Victora et al. (2015) - Findings

Observed a dose-response association between breastfeeding duration for IQ and educational attainment.

Those breastfed for ≥ 12m have:

• 3.76 higher IQ points advantage

• 0.91 more years of education

• Higher monthly incomes (341 Brazilian reals)

Than those breastfed for <1 month.

Riskin et al. (2011)

• Breastmilk=personalised medicine

• Changes in immunomodulatory constituents of human milk in response to active infection in the nursing infant

• A larger no. of CD45 and macrophages in breastmilk correlated with children not being unwell

Powls & Cooke (1988)

Median IQ lower in adolescence of low birthweight children than the normal population mean (100)

Effect of general anaesthesia in early childhood

• Animal findings - raised concerns (Roddy, 2012)

• Biological siblings - no adverse effect (O’Leary et al., 2018)

• Glatz et al. (2017) found exposure to anaesthesia before age 4 was associated with a 0.97% lower IQ score

  • 33,514 children relative to 159619 control children; IQ age 18, Swedish military conscription – a very small effect.

Effects of chemotherapy in early childhood

• 10-year literature review

• Concerns raised about neurocognitive late effects after chemotherapy.

• Resulted in deficits in cognitive abilities and IQ

• (Anderson & Kunin-Bateson, 2009)

Effect of Early Childhood Blood Lead Levels - Chatham-Stephens et al. (2010)

Noncancer effects of lead

• No evidence for mild mental retardation

• Significant evidence for decreased IQ

Effect of Early Childhood Blood Lead Levels (BLL) - Boyle et al. (2021)

• BLL (6-24m) → Decrease in IQ points (5-10 years)

• Black children → Highest BLL

• BLL (12-24m) → Black infants 46-55% greater estimated loss of IQ points than white or Hispanic

• Lead accounted for 74% of financial burden ($554 billion) due to loss in IQ points

The Flynn Effect

• Rise in average IQ scores over time across many countries and populations

Flynn (1987)

• IQ scores in Britain, Netherlands, Israel, Norway, Belgium increased over time (1942→1992)

• Gains are greatest for fluid intelligence

The Flynn Effect - Country Development and Age

Wonggupparaj et al. (2015)

• Raven Progressive Matrices

• Steeper gains in IQ are evident in developing countries. 

  • More stable gains in IQ in developed

• Evident across all age groups

The Anti-Flynn Effect

Norwegian IQ study - Sundet (2004)

• Significant decline

• The early, large gains are not evident at the more recent time points

• There is evidence of a decline in IQ across these latter generations.

Reversal confirmed in subsequent studies.

The Anti-Flynn effect in the US (2006-2018)

Dworak (2023)

• ICAR scores (cognitive ability) gradually declined across all levels of education

Interpreting the Flynn effect

• Not easily explained by changes in knowledge transmission or teaching, as greatest in fluid intelligence.

• It could be to do with practices or the valuation of abstract reasoning.

  • Flynn: fluid IQ tests measure something only loosely related to what we would call ‘intelligence’ in actual life

• Material explanations supported by close associations with height, nutrition etc.

Causes of The Flynn Affect - Kenya Study

Daley (2003)

• Increase in intelligence between 1984-1998

• Aligned with caloric and protein intake increase and hookworm decrease

• However, haemoglobin deficiency decreased

Causes of The Flynn Affect - Norwegian study

• There is a close mirroring between changes in general ability scores and changes in height.

Interpreting the Flynn effect

The Flynn effect was a transient phenomenon that reflected a boost in IQ driven by environmental factors, but that was masking:

• An underlying dysgenic trend (more intelligent people having fewer children)

• Compositional change from immigration

Hence, “the anti-Flynn effect being attributed mainly to genetics and immigration” (Riedermann, Becker & Coyle, 2017).

How to assses the Anti-flynn effect

• If caused by:

  • Dysgenic trends

  • Compositional change from immigration

• Then the underlying causal factors:

  • Will NOT be operating within families.

Bratsberg & Rogeberg (2018) - Method

• Modelled within-family variation within the Norwegian dataset.

• Population-covering data registers of IQ scores from Norway across 30 birth cohorts.

• Data registers contain information to precisely identify:

  • family relationships, birth order, and siblings

• without ability scores from military conscription testing.

Bratsberg & Rogeberg (2018) - Findings

• Found that within-family variation fully recovered both the timing and the magnitude of the increase and decline in the Flynn effect.

• Hence, the authors effectively disproved the dysgenic fertility and compositional change from immigration hypotheses.

• The causal factors of the Flynn effect must be environmental.

Individual differences in intelligence

Intelligence is a quality of

• Development (influenced by our DNA, our environments, and the interaction between the two).

• Health across the life span [G, E, and GxE]

Wechsler Intelligence Scale for Children (WISC)

Modified WAIS

• A standardised, individually administered intelligence test designed to assess the cognitive abilities of children and adolescents

Leiter-3 Test

• A nonverbal test of intelligence and cognitive abilities

• Designed for ages 3-70+

Contributions to Learning - Deary et al. (2017)

IQ and academic achievement are strongly related

Contributions to Learning - Ceci & Williams (1997)

Educational attainment is a stronger predictor than IQ of occupational success and income

Contributions to Learning - Kriegbaum et al. (2018)

Moderate correlation between intelligence and achievement

Contributions to Learning - Bergold & Steinmayr (2018)

Relationship between intelligence and academic ability is stronger when students score

• Higher in conscientiousness

• Lower in neuroticism

Contributions to Learning - Steinmayr et al. (2010)

Relationship between GPA and intelligence is moderated by quality of performance

Heaven & Ciarrochi (2012)

Intellect is associated with higher academic ability in those with high ability but not low ability

Contributions to learning - Primi et al. (2010)

Higher levels of intelligence associated with steeper improvement curves in 11-14 year olds in maths

Contributions to learning - PISA; Lynn & Mikk (2009)

Significant positive correlations between IQ and reading comprehension, maths ability, and science understanding

WM vs intelligence

• Correlations less than unity between constructs (Acherman et al., 2005)

• Simple short term storage component accounts for relationship between WM and g (Colom et al., 2008)

• Focus of attention drives relationship between WM and g (Gray et al., 2017)

• WM and fluid intelligence contribute to reading speed and comprehension (Johann et al., 2020)

WM vs Intelligence - Giofrè et al. (2013; 2017)

• WM predicts 65% of variance of g in 4^th and 5^th grade children

• G better predictor of achievement than WM in 6th and 8th grade children

Alloway & Alloway (2010)

WM at 5 was best predictor of literacy and numeracy 6 years later