Previous results have been mixed regarding the role of the apolipoprotein E e4 (APOE e4) allele in later-life depression: some studies note that carriers experience greater symptoms and increased risk while others find no such association. However, there are few prospective, population-based studies of the APOE e4-depression association and fewer that examine depressive symptom trajectory and depression risk longitudinally. We examined the association between APOE e4 allele status and longitudinal change in depressive symptoms and depression risk in later-life, over a 12-year follow-up period.

We used data from 690 participants of the Lothian Birth Cohort 1936 who took part in the Scottish Mental Survey 1947 (aged 11) and were followed-up in later-life over five waves from 2004 to 2019 (aged 70–82). We used APOE e4 allele status to predict longitudinal change in depressive symptom scores and risk of depression (defined by a symptom score threshold or use of depression-related medication). Models were adjusted for sex, childhood cognitive ability, childhood social class, education, adult social class, smoking status and functional limitations at baseline.

Depressive symptom scores increased with age. Once adjusted for covariates, APOE e4 allele status did not significantly predict symptom score trajectories or depression risk. Greater functional limitations at baseline significantly predicted poorer symptom score trajectories and increased depression risk (defined by medications). APOE e4 allele status did not significantly moderate the contribution of sex, education or functional limitations.

There was no evidence that APOE e4 carriers experience an increased risk for later-life depression.

Loneliness is a growing public health issue in the developed world. Among older adults, loneliness is a particular challenge, as the older segment of the population is growing and loneliness is comorbid with many mental as well as physical health issues. Comorbidity and common cause factors make identifying the antecedents of loneliness difficult, however, contemporary machine learning techniques are positioned to tackle this problem.

This study analyzed four cohorts of older individuals, split into two age groups – 45–69 and 70–79 – to examine which common psychological and sociodemographic are associated with loneliness at different ages. Gradient boosted modeling, a machine learning technique, and regression models were used to identify and replicate associations with loneliness.

In all cohorts, higher emotional stability was associated with lower loneliness. In the older group, social circumstances such as living alone were also associated with higher loneliness. In the younger group, extraversion's association with lower loneliness was the only other confirmed relationship.

Different individual and social factors might underlie loneliness differences in distinct age groups. Machine learning methods have the potential to unveil novel associations between psychological and social variables, particularly interactions, and mental health outcomes.

Major depressive disorder and neuroticism (Neu) share a large genetic basis. We sought to determine whether this shared basis could be decomposed to identify genetic factors that are specific to depression.

We analysed summary statistics from genome-wide association studies (GWAS) of depression (from the Psychiatric Genomics Consortium, 23andMe and UK Biobank) and compared them with GWAS of Neu (from UK Biobank). First, we used a pairwise GWAS analysis to classify variants as associated with only depression, with only Neu or with both. Second, we estimated partial genetic correlations to test whether the depression's genetic link with other phenotypes was explained by shared overlap with Neu.

We found evidence that most genomic regions (25/37) associated with depression are likely to be shared with Neu. The overlapping common genetic variance of depression and Neu was genetically correlated primarily with psychiatric disorders. We found that the genetic contributions to depression, that were not shared with Neu, were positively correlated with metabolic phenotypes and cardiovascular disease, and negatively correlated with the personality trait conscientiousness. After removing shared genetic overlap with Neu, depression still had a specific association with schizophrenia, bipolar disorder, coronary artery disease and age of first birth. Independent of depression, Neu had specific genetic correlates in ulcerative colitis, pubertal growth, anorexia and education.

Our findings demonstrate that, while genetic risk factors for depression are largely shared with Neu, there are also non-Neu-related features of depression that may be useful for further patient or phenotypic stratification.

Cognitive impairment is strongly linked with persistent disability in people with mood disorders, but the factors that explain cognitive impairment in this population are unclear.

To estimate the total effect of (a) bipolar disorder and (b) major depression on cognitive function, and the magnitude of the effect that is explained by potentially modifiable intermediate factors.

Cross-sectional study using baseline data from the UK Biobank cohort. Participants were categorised as having bipolar disorder (n = 2709), major depression (n = 50 975) or no mood disorder (n = 102 931 and n = 105 284). The outcomes were computerised tests of reasoning, reaction time and memory. The potential mediators were cardiometabolic disease and psychotropic medication. Analyses were informed by graphical methods and controlled for confounding using regression, propensity score-based methods and G-computation.

Group differences of small magnitude were found on a visuospatial memory test. Z-score differences for the bipolar disorder group were in the range −0.23 to −0.17 (95% CI −0.39 to −0.03) across different estimation methods, and for the major depression group they were approximately −0.07 (95% CI −0.10 to −0.03). One-quarter of the effect was mediated via psychotropic medication in the bipolar disorder group (−0.05; 95% CI −0.09 to −0.01). No evidence was found for mediation via cardiometabolic disease.

In a large community-based sample in middle to early old age, bipolar disorder and depression were associated with lower visuospatial memory performance, in part potentially due to psychotropic medication use. Mood disorders and their treatments will have increasing importance for population cognitive health as the proportion of older adults continues to grow.

I.J.D. is a UK Biobank participant. J.P.P. is a member of the UK Biobank Steering Committee.

Substantial clinical heterogeneity of major depressive disorder (MDD) suggests it may group together individuals with diverse aetiologies. Identifying distinct subtypes should lead to more effective diagnosis and treatment, while providing more useful targets for further research. Genetic and clinical overlap between MDD and schizophrenia (SCZ) suggests an MDD subtype may share underlying mechanisms with SCZ.

The present study investigated whether a neurobiologically distinct subtype of MDD could be identified by SCZ polygenic risk score (PRS). We explored interactive effects between SCZ PRS and MDD case/control status on a range of cortical, subcortical and white matter metrics among 2370 male and 2574 female UK Biobank participants.

There was a significant SCZ PRS by MDD interaction for rostral anterior cingulate cortex (RACC) thickness (β = 0.191, q = 0.043). This was driven by a positive association between SCZ PRS and RACC thickness among MDD cases (β = 0.098, p = 0.026), compared to a negative association among controls (β = −0.087, p = 0.002). MDD cases with low SCZ PRS showed thinner RACC, although the opposite difference for high-SCZ-PRS cases was not significant. There were nominal interactions for other brain metrics, but none remained significant after correcting for multiple comparisons.

Our significant results indicate that MDD case-control differences in RACC thickness vary as a function of SCZ PRS. Although this was not the case for most other brain measures assessed, our specific findings still provide some further evidence that MDD in the presence of high genetic risk for SCZ is subtly neurobiologically distinct from MDD in general.

Environmental risk factors for dementia are poorly understood. Aluminium and fluorine in drinking water have been linked with dementia but uncertainties remain about this relationship.

In the largest longitudinal study in this context, we set out to explore the individual effect of aluminium and fluoride in drinking water on dementia risk and, as fluorine can increase absorption of aluminium, we also examine any synergistic influence on dementia.

We used Cox models to investigate the association between mean aluminium and fluoride levels in drinking water at their residential location (collected 2005–2012 by the Drinking Water Quality Regulator for Scotland) with dementia in members of the Scottish Mental Survey 1932 cohort who were alive in 2005.

A total of 1972 out of 6990 individuals developed dementia by the linkage date in 2012. Dementia risk was raised with increasing mean aluminium levels in women (hazard ratio per s.d. increase 1.09, 95% CI 1.03–1.15, P < 0.001) and men (1.12, 95% CI 1.03–1.21, P = 0.004). A dose-response pattern of association was observed between mean fluoride levels and dementia in women (1.34, 95% CI 1.28–1.41, P < 0.001) and men (1.30, 95% CI 1.22–1.39, P < 0.001), with dementia risk more than doubled in the highest quartile compared with the lowest. There was no statistical interaction between aluminium and fluoride levels in relation with dementia.

Higher levels of aluminium and fluoride were related to dementia risk in a population of men and women who consumed relatively low drinking-water levels of both.

Low birth weight has been inconsistently associated with risk of developing affective disorders, including major depressive disorder (MDD). To date, studies investigating possible associations between birth weight and bipolar disorder (BD), or personality traits known to predispose to affective disorders such as neuroticism, have not been conducted in large cohorts.

To assess whether very low birth weight (<1500 g) and low birth weight (1500–2490 g) were associated with higher neuroticism scores assessed in middle age, and lifetime history of either MDD or BD. We controlled for possible confounding factors.

Retrospective cohort study using baseline data on the 83 545 UK Biobank participants with detailed mental health and birth weight data. Main outcomes were prevalent MDD and BD, and neuroticism assessed using the Eysenck Personality Inventory Neuroticism scale - Revised (EPIN-R)

Referent to normal birth weight, very low/low birth weight were associated with higher neuroticism scores, increased MDD and BD. The associations between birth weight category and MDD were partially mediated by higher neuroticism.

These findings suggest that intrauterine programming may play a role in lifetime vulnerability to affective disorders.

The relative contribution of demographic, lifestyle and medication factors to the association between affective disorders and cardiometabolic diseases is poorly understood.

To assess the relationship between cardiometabolic disease and features of depresion and bipolar disorder within a large population sample.

Cross-sectional study of 145 991 UK Biobank participants: multivariate analyses of associations between features of depression or bipolar disorder and five cardiometabolic outcomes, adjusting for confounding factors.

There were significant associations between mood disorder features and ‘any cardiovascular disease’ (depression odds ratio (OR) = 1.15, 95% CI 1.12–1.19; bipolar OR = 1.28, 95% CI 1.14–1.43) and with hypertension (depression OR = 1.15, 95% CI 1.13–1.18; bipolar OR = 1.26, 95% CI 1.12–1.42). Individuals with features of mood disorder taking psychotropic medication were significantly more likely than controls not on psychotropics to report myocardial infarction (depression OR = 1.47, 95% CI 1.24–1.73; bipolar OR = 2.23, 95% CI 1.53–3.57) and stroke (depression OR = 2.46, 95% CI 2.10–2.80; bipolar OR = 2.31, 95% CI 1.39–3.85).

Associations between features of depression or bipolar disorder and cardiovascular disease outcomes were statistically independent of demographic, lifestyle and medication confounders. Psychotropic medication may also be a risk factor for cardiometabolic disease in individuals without a clear history of mood disorder.

We examined the associations between serum cholesterol measures, statin use, and cognitive function measured in childhood and in old age. The possibility that lifelong (trait) cognitive ability accounts for any cross-sectional associations between cholesterol and cognitive performance in older age, seen in observational studies, has not been tested to date.

Participants were 1,043 men and women from the Lothian Birth Cohort 1936 Study, most of whom had participated in a nationwide IQ-type test in childhood (Scottish Mental Survey of 1947), and were followed up at about age 70 years. Serum cholesterol measures included total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), triglycerides, and cholesterol:HDL cholesterol ratio. Cognitive outcome measures were age 70 IQ (using the same test as at age 11 years), general cognitive ability (g), processing speed, memory, and verbal ability.

Higher TC, higher HDL-C, and lower triglycerides were associated with higher age 70 cognitive scores in most cognitive domains. These relationships were no longer significant after covarying for childhood IQ, with the exception a markedly attenuated association between TC and processing speed, and triglycerides and age 70 IQ. In the fully adjusted model, all conventionally significant (p < 0.05) effects were removed. Childhood IQ predicted statin use in old age. Statin users had lower g, processing speed, and verbal ability scores at age 70 years after covarying for childhood IQ, but significance was lost after adjusting for TC levels.

These results suggest that serum cholesterol and cognitive function are associated in older age via the lifelong stable trait of intelligence. Potential mechanisms, including lifestyle factors, are discussed.

Reports of attitudes to aging from older people themselves are scarce. Which life course factors predict differences in these attitudes is unknown.

We investigated life course influences on attitudes to aging in healthy, community-dwelling people in the UK. Participants in the Lothian Birth Cohort 1936 completed a self-report questionnaire (Attitudes to Aging Questionnaire, AAQ) at around age 75 (n = 792, 51.4% male). Demographic, social, physical, cognitive, and personality/mood predictors were assessed, around age 70. Cognitive ability data were available at age 11.

Generally positive attitudes were reported in all three domains: low Psychosocial Loss, high Physical Change, and high Psychological Growth. Hierarchical multiple regression found that demographic, cognitive, and physical variables each explained a relatively small proportion of the variance in attitudes to aging, with the addition of personality/mood variables contributing most significantly. Predictors of attitudes to Psychosocial Loss were high neuroticism; low extraversion, openness, agreeableness, and conscientiousness; high anxiety and depression; and more physical disability. Predictors of attitudes to Physical Change were: high extraversion, openness, agreeableness, and conscientiousness; female sex; social class; and less physical disability. Personality predictors of attitudes to Psychological Growth were similar. In contrast, less affluent environment, living alone, lower vocabulary scores, and slower walking speed predicted more positive attitudes in this domain.

Older people's attitudes to aging are generally positive. The main predictors of attitude are personality traits. Influencing social circumstances, physical well-being, or mood may result in more positive attitudes. Alternatively, interventions to influence attitudes may have a positive impact on associated physical and affective changes.

Evidence from observational studies to date suggests that healthy dietary patterns are associated with better cognitive performance in later life. We examined the extent to which childhood intelligence quotient (IQ) and socioeconomic status account for this association.

Analyses were carried out on 882 participants in the Lothian Birth Cohort 1936 Study. Four dietary patterns were extracted using principal components analysis of a food frequency questionnaire, namely “Mediterranean-style,” “health aware,” “traditional,” and “sweet foods.” Cognitive function was assessed at the age of 70 years, including general (g) cognitive ability, processing speed, memory, and verbal ability.

Before adjustment for childhood IQ and socioeconomic status, the “Mediterranean-style” dietary pattern was associated with significantly better cognitive performance (effect size as partial eta-square (ηp2) range = 0.005 to 0.055), and the “traditional” dietary pattern was associated with poorer performance on all cognitive domains measured in old age (ηp2 = 0.009 to 0.103). After adjustment for childhood IQ (measured at the age of 11 years) and socioeconomic status, statistical significance was lost for most associations, with the exception of verbal ability and the “Mediterranean-style” pattern (National Adult Reading Test (NART) ηp2 = 0.006 and Wechsler Test of Adult Reading (WTAR) ηp2 = 0.013), and the “traditional” pattern (NART ηp2 = 0.035 and WTAR ηp2 = 0.027).

Our results suggest a pattern of reverse causation or confounding; a higher childhood cognitive ability (and adult socioeconomic status) predicts adherence to a “healthy” diet and better cognitive performance in old age. Our models show no direct link between diet and cognitive performance in old age; instead they are related via the lifelong-stable trait of intelligence.