LONGVIEW: COGNITIVE CAPITAL OVER THE LAST 50 YEARS
SEMINAR 5 ADULT LIFE CHANCES: HEALTH
Based on work led byDavid Batty, Catherine Gale, Stephani Hatch,
Barbara Jefferis, Diana Kuh, and Marcus Richards
PATHWAYS FROM COGNITIVE CAPITAL TO HEALTH
Mediation
Cognition may mediate early circumstances that influence health:
• material home conditions (wealth)• family interaction (psychosocial)• developing health itself
PATHWAYS FROM COGNITIVE CAPITAL TO HEALTH
Social causation
Cognition can determine the acquisition of factors that influence health:
• material resources (wealth)• psychosocial capital (status, control, wellbeing) • self-care/health literacy (lifestyle, health service use)
PATHWAYS FROM COGNITIVE CAPITAL TO HEALTH
Cognition as a biomarker
Cognition may mark underlying physiological processesthat regulate health:
• central nervous system• autonomic nervous system• endocrine axes (growth, thyroid, HPA, HPG)• oxidative biochemistry• immune function• genetic pleiotropy
British 1946 birth cohort: cumulative male death rate 9 to 54 years for the highest and lowest quarters of the cognitive score at age 8 years.
0.88
0.9
0.92
0.94
0.96
0.98
1
108 144 180 216 252 288 324 360 396 432 468 504 540 576 612 648
Age in years (and months since birth)
Prop
ortio
n al
ive
9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54
Highest quartile
Lowest quartile
Kuh, Richards, Hardy, Butterworth & Wadsworth Int J Epidemiol 2004; 33: 408-413
Cognitive ability tests____________________________________________________
1946 cohort 1970 cohort____________________________________________________
Non-Verbal intelligence Non-verbal (BAS)Verbal intelligence - Recall of digitsVocabulary - MatricesPronunciation Verbal (BAS)Arithmetic - Word definitions
- Word similarities___________________________________________________________
LIFESTYLE:
HEALTH-RELATED BEHAVIOURS
Ever smoked cigarettes___________________________________________________________
1946 cohort (36 y) 1970 cohort (30 y) (63.8%) (55.4%)
___________________________________________________________
Unadjusted 0.83 (0.76, 0.92) p < 0.001 -----
Gender 0.84 (0.76, 0.92) p < 0.001 0.92 (0.88, 0.96)
Parental SC 0.86 (0.77, 0.95) p = 0.003 0.93 (0.89, 0.97)
Own education 1.07 (0.95, 1.22) p = 0.26 1.08 (1.03, 1.14)
Adult SC 0.90 (0.80, 1.00) p = 0.05 1.01 (0.96, 1.06)
Adult earnings 0.80 (0.72, 0.88) p < 0.001 0.93 (0.88, 0.98)
All 1.09 (0.96, 1.24) p = 0.19 1.08 (1.02, 1.15)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. In press: Journal of Epidemiology and Community Health
Quit smoking cigarettes___________________________________________________________
1946 cohort (36 y) 1970 cohort (30 y) (33.4%) (34.2%)
___________________________________________________________
Unadjusted 1.31 (1.16, 1.48) p < 0.001 -----
Gender 1.31 (1.16, 1.48) p < 0.001 1.25 (1.18, 1.34)
Parental SC 1.29 (1.13, 1.47) p < 0.001 1.18 (1.10, 1.26)
Own education 1.13 (0.97, 1.32) p = 0.11 1.10 (1.02, 1.18)
Adult SC 1.22 (1.06, 1.40) p = 0.005 1.15 (1.07, 1.23)
Adult earnings 1.30 (1.15, 1.48) p < 0.001 1.16 (1.07, 1.25)
All 1.11 (0.95, 1.31) p = 0.19 1.02 (0.93, 1.11)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. In press: Journal of Epidemiology and Community Health
1946 cohort: odds of being CAGE positive at 53 yearsper point increase in childhood cognition at 8 years
Odds = 1.04 (1.01 – 1.07) p = 0.02
Adjusting for:
• Gender• SES origin and early adversity (father’s social class, mother’s education, parental
divorce, suboptimal maternal management and poor material home conditions) • Highest educational qualifications by 26 years)• Adult SES (social class and household income)• Adult stressors (unemployment and economic hardship)• Adult health behaviours (exercise and smoking)
Hatch, Jones, Kuh, Hardy, Wadsworth & Richards Soc Sci Med 2007; 64: 2285-2296
Ever CAGE positive (men)___________________________________________________________
1946 cohort (43 y) 1970 cohort (30 y) (13.7%) (20.4%)
___________________________________________________________
Unadjusted 0.99 (0.84, 1.18) p = 0.93 -----
Parental SC 1.00 (0.83, 1.21) p = 0.98 1.13 (1.04, 1.24)
Own education 1.03 (0.82, 1.29) p = 0.79 1.13 (1.04, 1.24)
Adult SC 1.00 (0.82, 1.22) p = 0.99 1.18 (1.07, 1.30)
Adult earnings 0.99 (0.82, 1.19) p = 0.90 1.15 (1.03, 1.26)
Has children 0.99 (0.83, 1.18) p = 0.92 1.11 (1.02, 1.22)
All 1.03 (0.81, 1.30) p = 0.82 1.16 (1.04, 1.28)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. In press: American Journal of Public Health
Ever CAGE positive (women)___________________________________________________________
1946 cohort (43 y) 1970 cohort (30 y) (9.3%) (8.1%)
___________________________________________________________
Unadjusted 1.43 (1.06, 1.94) p = 0.02 1.44 (1.24, 1.67)
Parental SC 1.37 (0.99, 1.91) p = 0.06 1.42 (1.22, 1.65)
Own education 1.09 (0.74, 1.62) p = 0.67 1.46 (1.24, 1.71)
Adult SC 1.24 (0.89, 1.74) p = 0.20 1.35 (1.16, 1.59)
Adult earnings 1.16 (0.84, 1.59) p = 0.37 1.44 (1.24, 1.67)
Has children 1.41 (1.04, 1.91) p = 0.03 1.36 (1.18, 1.58)
All 1.02 (0.68, 1.53) p = 0.92 1.38 (1.17, 1.64)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. In press: American Journal of Public Health
Any physical exercise___________________________________________________________
1946 cohort (36 y) 1970 cohort (30 y) (65.4%)
___________________________________________________________
Unadjusted 1.40 (1.28, 1.54) p < 0.001 1.21 (1.14, 1.27)
Gender 1.42 (1.29, 1.56) p < 0.001 1.20 (1.14, 1.27)
Parental SC 1.35 (1.22, 1.49) p < 0.001 1.16 (1.10, 1.23)
Own education 1.23 (1.09, 1.39) p = 0.001 1.06 (1.00, 1.12)
Adult SC 1.26 (1.13, 1.40) p < 0.001 1.15 (1.09, 1.22)
Adult earnings 1.30 (1.18, 1.43) p < 0.001 1.16 (1.09, 1.24)
All 1.19 (1.05, 1.35) p = 0.006 1.02 (0.94, 1.09)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. Pediatrics 2007 (www.pediatrics.org)
Intense physical exercise___________________________________________________________
1946 cohort (53 y) 1970 cohort (30 y) (37%)
___________________________________________________________
Unadjusted 1.80 (1.60, 2.02) p < 0.001 1.35 (1.29, 1.41)
Gender 1.81 (1.61, 2.04) p < 0.001 1.30 (1.24, 1.36)
Parental SC 1.71 (1.51, 1.94) p < 0.001 1.31 (1.25, 1.37)
Own education 1.43 (1.23, 1.65) p < 0.001 1.25 (1.19, 1.31)
Adult SC 1.60 (1.40, 1.82) p < 0.001 1.22 (1.16, 1.28)
Adult earnings 1.69 (1.50, 1.90) p < 0.001 1.20 (1.14, 1.27)
All 1.39 (1.19, 1.62) p < 0.001 1.15 (1.08, 1.22)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. Pediatrics 2007 (www.pediatrics.org)
Healthy food choice at 53 years (1946 cohort)
_________________________________________________
Unadjusted 1.53 (1.35, 1.73) p < 0.001
Gender 1.52 (1.35, 1.72) p < 0.001
Parental SC 1.46 (1.28, 1.66) p < 0.001
Own education 1.14 (0.98, 1.34) p = 0.09
Adult SC 1.44 (1.25, 1.65) p < 0.001
Adult earnings 1.57 (1.38, 1.78) p < 0.001
All 1.14 (0.97, 1.34) p = 0.11
_________________________________________________
Based on forthcoming work in collaboration with the MRC Human Nutrition
Research Centre, Elsie Widdowson Laboratories, Cambridge
Healthy food choice representing 1. frequency of breakfast, 2. type of milk,3. type of bread, 4. proportion of energy derived from fat, and 5. number ofdaily portions of fruit and vegetables (odds of > median total score)
Odds of being a vegetarian at 30 years (1970 cohort)
_______________________
Unadjusted 1.38 (1.24, 1.53)
Gender 1.42 (1.28, 1.59)
Parental SC 1.35 (1.21, 1.51)
Own education 1.16 (1.03, 1.30)
Adult SC 1.29 (1.15, 1.45)
All 1.20 (1.06, 1.36)
_______________________
Gale C, Deary I, Schoon I, Batty D. British Medical Journal 2007;334:245.
Health behaviours: summary
• Childhood IQ is mainly protective of harmful health- related behaviours, although it appears to be associated with increased risk of potential alcohol abuse
• These effects were often explained by educational attainment or adult SES, although there were independent effects of IQ on on risk of alcohol abuse and being a vegetarian in the 1970 and on likelihood of undertaking physical exercise in both cohorts
• There were few obvious cohort effects, although positive associations between IQ and potential alcohol abuse appeared to be stronger in the 1958 cohort
CHRONIC PHYSICAL DISEASE
Hypertension (self reported)___________________________________________________________
1946 cohort (36 y) 1970 cohort (30 y) (3.1%) (7.7%)
___________________________________________________________
Unadjusted 0.75 (0.59, 0.95) p = 0.02 -----
Gender 0.74 (0.58, 0.95) p = 0.02 0.90 (0.83, 0.98)
Parental SC 0.75 (0.57, 0.97) p = 0.03 0.92 (0.84, 0.99)
Own education 0.69 (0.51, 0.95) p = 0.02 0.93 (0.85, 1.01)
Adult SC 0.77 (0.58, 1.02) p = 0.07 0.91 (0.83, 0.99)
Adult earnings 0.78 (0.61, 1.01) p = 0.06 0.89 (0.80, 0.98)
All 0.70 (0.51, 0.98) p = 0.04 0.94 (0.84, 1.06)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. In press: Journal of Epidemiology and Community Health
Obesity (BMI > 30 kg/m2 : WHO)___________________________________________________________
1946 cohort (36 y) 1970 cohort (30 y) (5.6%) (11.3%)
___________________________________________________________
Unadjusted 0.73 (0.61, 0.88) p = 0.001 -----
Gender 0.73 (0.61, 0.88) p = 0.001 0.84 (0.79, 0.91)
Parental SC 0.78 (0.94, 0.95) p = 0.02 0.89 (0.83, 0.96)
Own education 0.79 (0.62, 1.00) p = 0.05 0.93 (0.86, 1.00)
Adult SC 0.78 (0.63, 0.97) p = 0.02 0.88 (0.81, 0.95)
Adult earnings 0.76 (0.63, 0.92) p = 0.005 0.86 (0.79, 0.94)
All 0.82 (0.64, 1.05) p = 0.12 0.97 (0.88, 1.07)
____________________________________________________________
Batty D, Deary I, Schoon I, Gale C. In press: Journal of Epidemiology and Community Health
Metabolic syndrome at 53 years (1946 cohort)
_________________________________________________
Unadjusted 0.82 (0.72, 0.94) p = 0.004
Gender 0.82 (0.72, 0.94) p = 0.004
Parental SC 0.86 (0.74, 0.99) p = 0.04
Own education 0.96 (0.80, 1.14) p = 0.64
Adult SC 0.85 (0.73, 1.00) p = 0.04
Adult earnings 0.79 (0.69, 0.91) p = 0.001
All 0.98 (0.82, 1.18) p = 0.84
_________________________________________________
Metabolic syndrome identified in the 1946 cohort by Claudia Langenberg(American Journal of Public Health 2006; 96: 2216-2221)
National Cholesterol Education Program Adult Treatment Panel III (ATPIII) definition (21.9%)
TIMING OF THE NATURAL MENOPAUSE
• Mediation: Cognition may mediate early circumstances that influence reproductive ageing
• Social causation: Education and cognition determine health behaviours (e.g. smoking) that trigger early menopause
• Cognition as a biomarker: oestrogen facilitates neural growth, prevents neuronal damage and death, and increases cerebral blood flow
Survivor functions for age at menopause by cognitive ability score at age 8 years
0.00
0.25
0.50
0.75
1.00
34 36 38 40 42 44 46 48 50 52 54 56 58
Age (years)
Lowest third Highest third
Pro
po
r tio
n p
r e/p
eri
men
op
ausa
l
Richards et al. Neurology 1999; Kuh et al. Menopause 2005
Effect of one standard deviation change in cognitive ability at 8 years on age at menopause
Hazard Ratio (95%CI)
Unadjusted 0.89 (0.82,0.96)
Adjusted for been breastfed 0.88 (0.81,0.96)
Adjusted for father’s social class 0.91 (0.83,0.99)
Adjusted for parental divorce (by 15) 0.89 (0.82,0.96)
Adjusted for parity 0.88 (0.80,0.95)
Adjusted for smoking 0.91 (0.83,0.99)
Fully adjusted 0.90 (0.82,0.98)
Kuh et al. Menopause 2005; 12: 475-82
COX PROPORTIONAL HAZARDS FORCOGNITION AND AGE AT MENOPAUSE
___________________________________________________________
Age of test Hazard ratio (95% CI) p value
___________________________________________________________
8 years 0.73 (0.62 - 0.86) <0.001
11 years 0.78 (0.67 - 0.91) 0.002
15 years 0.80 (0.67 - 0.94) 0.007
26 years 0.85 (0.72 - 1.00) 0.048
__________________________________________________________________
Chronic physical disease: summary
• Childhood IQ is associated with reduced risk of chronic physical disease
• As with health behaviours, this was mostly explained by educational attainment or adult SES, although there is an independent inverse association between IQ and risk of early natural menopause, possibly reflecting early programming of the reproductive hormonal axis
• There were few obvious cohort effects, although associations between IQ and self-reported hypertension appeared to be stronger in the 1946 cohort
MENTAL HEALTH
1946 cohort: total GHQ-28 score at 53 yearsper point increase in childhood cognition at 8 years
Regression coefficient (men) = -0.04 (-0.22, -0.15) p = 0.70Regression coefficient (women) = -0.29 (-0.52, -0.06) p = 0.01
Adjusting for:
• SES origin and early adversity (father’s social class, mother’s education, parental divorce, suboptimal maternal management
and poor material home conditions) • Highest educational qualifications by 26 years)• Adult SES (social class and household income)• Adult stressors (unemployment and economic hardship)• Adult health behaviours (exercise and smoking)
Hatch, Jones, Kuh, Hardy, Wadsworth & Richards Soc Sci Med 2007; 64: 2285-2296
LITERACY, COGNITIVE RESERVE,AND COGNITIVE DECLINE
Father’s occupation
Cognitive abilityat 8 years
Education by 26 years
Own occupationat 43 years
literacyat 53 years
0.33 0.18
0.50
0.05
0.07
0.47
0.60
0.24
0.14
1946 cohort 1958 cohort
Father’s occupation
Cognitive abilityat 7 years
Education by 23 years
Own occupationat 33 years
literacyat 43 years
0.31
0.54
0.43
0.01
0.08
0.11
0.21
0.21
0.41
Richards, Power & Sacker (In press: J Epidemiol Community Health)
Priorability
Age
Cognitive decline (1946 cohort)
Richards M, Shipley B, Fuhrer F & Wadsworth MBritish Medical Journal 2005; 328: 552-554
Age specific prevalence of Alzheimer’s disease
0
5
10
15
20
25
30
0 65 70 75 80 85 90
Age (years)
Prevalence (%)
Clinical expressionof disease
Other influences ondisease expression:PersonalityHealth service delivery and uptakeCultural norms
BRAIN LESIONS
INFLUENCING FACTORS
Genes
Early social and material environment
Educational and occupational attainment
Physical health
Health behaviours and lifestyle
PREMORBIDBRAIN SIZE COGNITIVEAND FUNCTION ABILITY
(RESERVE)
Richards M & Deary I. A life course approach to cognitive reserve: a model for cognitive aging and development? Annals of Neurology 2005; 58: 617-620
But what about wisdom?
• The development of expertise
• The development of contextualisation
• Relativism of values and priorities
• Recognition and management of uncertainty