Intelligence, aging, life outcomes

How would you test stability of intelligence across a lifespan?

Scottish Mental Survey (1932 & 1947)

• Followed up with individuals who took part in survey, living in Edinburgh

• In 1998, 73 people retook the test they had taken at 11 yo

Perfect stability was not observed

• Scored better at 77 than 11

HOWEVER

• Those who scored well tended to score well again

• High correlation (0.6-0.7)

Development of cognitive abilities

Crystallised Abilities (Vocabulary) → Increase with age

Fluid Abilities (Processing Speed) → Decrease with age

Livingston et al. (2020)

• 40% of risk factors for dementia are potentially modifiable

• 60% of risk factors for dementia are currently unknown

Cognitive Reserve determined by 5 factors (Schoentgen, 2020)

• Environmental and Cognitive Enrichment in Childhood, such as:

  • Activities

  • Environment

  • Biology

  • Psychology

  • Education

• Act as Protective Factors in the Adult and Ageing Brain

Bililingualism and Cognitive reserve

• Can maintain cognitive functioning longer despite age-related brain changes or pathology.

• Later onset of dementia symptoms, even when brain imaging shows similar levels of neurodegeneration

• Influenced by factors such as age of acquisition, frequency of language use, and language proficiency

  • Lifelong biligualism → more protective effects

Deary et al. (2007) - Method

Intelligence and Educational Achievement

• 5-year prospective longitudinal study of 70,000+ English children

• Examined the association between:

  • IQ (11y)

  • Educational achievement/GCSE scores (16y) in 25 academic subjects

Deary et al. (2007) - Findings

• IQ (11y) was a powerful predictor of educational attainment at (16y)

• IQ accounted for a little over half of the differences in GCSE scores

• A considerable amount of the variation in GCSE scores = not related to CAT scores at age 11

• Implies IQ is a substantial component of GCSE success but not all

Spinath et al. (2006)

9y Children’s Maths, English and Science scores

• g = strongest predictor of overall educational achievement

• Self-perception of ability = strongest predictor of English and Maths

• Common variance in school achievement is explained by g and motivation.

Lu et al. (2010)

Cognitive variables to explain the variance in students’ Chinese and Maths scores

• WM → 17.8%

• Intelligence → 36.4%

Domain-specific motivational constructs only marginally predicted school achievement

Schmidt & Hunter (1998)

• Intelligence predicted 51% of the variance in job performance

• However, findings are more relevant for complex professional occupations

• Other relevant variables:

  • Integrity

  • Conscientiousness

  • Reference checks

  • Interview performance

• Also: measures of general intelligence are generally poor at measuring constructs such as creativity and wisdom

Job Performance - False claims about Emotional Intelligence (EI)

EI matters twice as much as IQ (Goleman, 1998)

90% of the difference between star performers and other workers is attributable to EI factors

85% of job success could be attributed to EI (Watkin, 2000).

What does Emotional intelligence (EI) predict?

Better social and work relationships

• Able to more accurately detect variations in personal heartbeat

  • Schneider, Lyons, & Williams (2005)

• Being more able to recognise and reason about emotional consequences (e.g. affective forecasting)

  • Dunn et al. (2007)

• Lower scores on measures of distress symptoms, even after controlling for variation attributable to variation in Big Five personality dimensions

  • David (2005)

Batty et al. (2009):

• Studied 1 million Swedish men; 20 years of follow-up (14,498 deaths)

• Every SD increase in IQ → 32% decrease in mortality risk 20 years later

Calvin et al. (2017)

Found an association between:

• Intelligence at age 11

  • Moray House test score

• Major causes of death at age 79

  • Age and sex adjusted hazard ratios and 95% confidence intervals

Ian Deary on Health and Intelligence

• Risk of dying is not low vs high IQ

• Positive correlation between IQ and life span

IQ and Mental Health

Lower childhood IQ is associated with an increased risk of:

• Schizophrenia (David et al., 1997, Osler et al., 2007, Gunnell et al., 2002, Zammit et al., 2004, Dickson et al., 2012),

• Post-traumatic Stress Disorder (PTSD) (Koenen et al., 2007, Kremen et al., 2007, Gale et al., 2008, Gale et al., 2010),

• Depression (Zammit et al., 2004, Gale et al., 2008).

Gale et al. (2008; 2010); Koenen et al. (2007); Osler et al. (2007)

There is an association between:

• SD increase in IQ score in youth

• A risk reduction (13-43 %) in mental health difficulties in adulthood

  • (Schizophrenia, PTSD and Depression)

Gale et al. (2009)

Milder psychological distress (Anxiety & Depression)

• Association between:

• SD increase in IQ scores at age 10/11

• Reduction in psychological distress

  • 39% at age 30 (in the 1958 cohort)

  • 23% at age 33 (in the 1970 cohort).

Iverson et al. (2023) - Method

Examined the association between childhood (age 11) cognitive ability and self-harm and suicide risk among a Scotland-wide cohort (N = 53,037)

Used hospital admission and mortality records to follow individuals from age 34→85

Adjusted for childhood and adulthood SES

Iverson et al. (2023) - Findings

Association between:

• Higher childhood intelligence

• Reduced risk of self-harm

Was NOT significantly associated with suicide risk

Ball et al. (2024) - Methods

Childhood intelligence and risk of depression in later-life

• Participants took part in the Scottish Mental Survey in 1947 (n = 70,805)

• Health data was acquired from electronic health records - hospital admissions and prescribing data - between 1980 and 2020 (n = 53,037)

• Analyses were adjusted for:

  • Childhood family structure (e.g. number of children in the family and birth order)

  • Adulthood SES and geographical factors

Ball et al. (2024) - Findings

• Association between:

  • Higher childhood intelligence

  • Reduced risk of depression in later life

• However, only when:

  • Identified using hospital admissions data

  • Not when inferred from prescribed drugs data

Non-exclusive hypotheses

Holistic view of human potential, where intelligence matters but does not wholly determine success or life outcomes.

• Multiple influences on life outcomes

• Interaction effects between intelligence and other factors

• Context of the environment can effect the expression of intelligence on real life outcomes

• Different people may achieve similar outcomes through different combinations of traits and circumstances.

What intelligence isn’t

• President Donald Trump has boasted about his performance on a screening test for mild cognitive impairment or early dementia

Montreal Cognitive Assessment (MOCA)

• Diagnostic test

  • Simple tasks such as naming or copying simple shapes

• Not a measurement of intelligence