Socioeconomic inequalities in lung and upper aero-digestive tract cancer incidence in Scotland

Sharpe, Katharine H. (2018) Socioeconomic inequalities in lung and upper aero-digestive tract cancer incidence in Scotland. PhD thesis, University of Glasgow.

Abstract

Socioeconomic inequality in cancer risk and incidence burden has received limited
attention compared to genetic and behavioural risk factors. Where they have been
studied, the temporal relationship between socioeconomic factors and cancer risk
has been under explored due to the mainly cross-sectional nature of most
research. Moreover, the inter-relationships of the multiple measures of
socioeconomic status and, in particular, area and individual measures and their
interaction with risk behaviours have also had limited attention. The overarching
aim of this thesis was to investigate socioeconomic inequalities in the risk of lung
and upper aero-digestive tract cancers and the relationship between this risk and
socioeconomic status, area and individual based measures of socioeconomic
circumstances, and behaviours such as smoking, alcohol consumption, diet and
exercise.

To understand and quantify the relative contribution by age, sex and tumour
subtype to the socioeconomic inequalities of all cancer risk, a descriptive
epidemiological study of cancer incidence in Scotland (2000-07) was undertaken.
Age standardised rates per 100,000 population were calculated by direct
standardisation to the European standard. A linear regression model was used to
calculate the Slope Index of Inequality (SII) and Relative Index of Inequality (RII)
which were employed to rank tumour and subtype contribution to all cancer risk
socioeconomic inequalities by age for each sex for lung and upper aero-digestive
tract (UADT) cancers separately. There were 216,305 cases excluding nonmelanoma
skin cancer (all cancer) comprising 37,274 lung, 8,216 head and neck
and 6,534 oesophageal cancers classified into anatomical or morphological
subtypes. Socioeconomic circumstances were measured using the Scottish Index of
Multiple Deprivation (SIMD). Analyses were partitioned by five-year age group and
sex. RII was adapted to rank the contribution of each tumour type to all cancer
socioeconomic inequalities and to examine subtype by age and sex simultaneously.
The rank was defined as the proportion of all cancer socioeconomic inequality.

All cancer socioeconomic inequality was greater for males than females (RII=0.366;
female RII=0.279). The combination of lung and UADT socioeconomic inequalities
contributed 91% and 81% respectively to all cancer socioeconomic inequality. For
both sexes lung and UADT subtypes showed significant socioeconomic inequalities
(P<0.001) except oesophageal adenocarcinoma in males (P=0.193); for females,
socioeconomic inequality was borderline significant (P=0.048). Although RII rank
differed by sex, all lung and larynx subtypes contributed the most to all cancer
socioeconomic inequality with RII rank for oral cavity, oesophagus–squamous cell
and oropharynx following. For males 40-44 years old, socioeconomic inequalities
increased abruptly peaking at 55-59 years. For females, socioeconomic inequalities
gradually peaked 10 years later. In both sexes, the socioeconomic inequalities peak
age preceded age of peak incidence. This study showed that socioeconomic
inequalities in lung and UADT cancers vary greatly by age, tumour subtype and sex;
these variations were likely to largely reflect differences between the sexes in risk
behaviours which vary by birth cohort and are socioeconomically patterned.

Longitudinal data enabled exploration of the temporal relationship between
socioeconomic status and cancer incidence. An investigation of several individual
and a single area-based measure of socioeconomic circumstances was undertaken
in the second study of this thesis. The effect of country of birth, marital status,
one area socioeconomic circumstances measure (Carstairs) and five individual
socioeconomic variables (economic activity, education, occupational social class,
car ownership, household tenure) on the risk associated with lung, UADT and all
cancer combined (excluding non-melanoma skin cancer) were explored. A linked
dataset using the Scottish Longitudinal Study and Scottish Cancer Registry was
created to follow 203,658 cohort members aged 15+ years from 1991-2006.
Relative risks (RR) were calculated using Poisson regression models by sex offset
for person-years of follow-up. There were 21,832 first primary tumours (including
3,505 lung and 1,206 UADT cancers). Regardless of cancer, economic inactivity
(versus activity) was associated with increased risk (male: RR 1.14 95% CI 1.10,
1.18; female: RR 1.06 95% CI 1.02, 1.11). For lung cancer, area deprivation
remained significant after full adjustment suggesting that the area deprivation
cannot be fully explained by individual variables. Not having a qualification (versus
degree) was associated with increased lung cancer risk; likewise for UADT cancer
risk (females only). Occupational social class associations were most pronounced
and elevated for UADT risk. No car access (versus ownership) was associated with
increased risk (excluding all cancer risk for males). Renting accommodation (versus
home ownership) was associated with increased lung cancer risk, UADT cancer risk
for males only and all cancer risk for females only. Regardless of cancer group,
elevated risk was associated with no education and living in deprived areas. This
study demonstrated that different and independent socioeconomic variables were
inversely associated (greater incidence with lower socioeconomic circumstances)
with different cancer risks in both sexes; no one socioeconomic variable had a
dominant risk association or captured all aspects of socioeconomic circumstances
or the full life-course. The association of multiple socioeconomic variables was
likely to reflect the complexity and multifaceted nature of low socioeconomic
circumstances as well as the various roles of these dimensions over the life-course.

A final study investigated the role of behaviours (smoking, alcohol, diet and
exercise) on the association of low socioeconomic circumstances with all cancer
risk and lung and upper aero-digestive tract cancers combined (LUADT). The
Scottish Cancer Registry and Scottish Health Survey data were linked to create a
population study (1995-2011). There were 42,983 adults over 16 years old who
were followed for 3,750,611 person-years. There were 2,130 first primary cancers
diagnosed including 453 LUADT cancers. Poisson regression models, minimally
adjusted by age and sex, were developed to estimate the risk association between
five individual socioeconomic variables (economic activity, highest qualification,
occupational social class, car ownership and housing tenure), one area-based
socioeconomic indicator (SIMD) and all cancer and LUADT cancer. A further
socioeconomic indicator was developed to reflect multiple low socioeconomic
circumstances. This was defined as the count, at the individual participant level,
of socioeconomic variables in the highest risk category. A similar multiple high risk
behaviour derived variable, defined as the count of highest risk category for the
following variables: current smoking status, units of alcohol consumed in a week,
daily fruit and vegetable consumption and exercise sessions per week, was also
calculated at the individual participant level. The minimally adjusted Poisson
models were successively adjusted for behaviours (smoking, alcohol, diet and
exercise) to establish any remaining contribution to cancer risk not explained by
behaviour. Multiple low socioeconomic circumstances were very strongly
associated with increased risk for both cancer groups. For all cancer risk, the
elevated risk was nearly fully attenuated for all categories of multiple low
socioeconomic circumstances when adjusted for smoking only. For LUADT cancer
and in the minimally adjusted model, the risk increased in a dose-response
manner. The risk associated with LUADT cancer for study participants in the
highest category of multiple low socioeconomic circumstances was more than
three-times greater when compared to their affluent counterparts (RR 3.35 95% CI
2.26, 4.97); this elevated risk remained at 86% compared to those with no
socioeconomic disadvantage, even after full adjustment for smoking, alcohol, diet
and exercise behaviours. When looking at single socioeconomic status (SES)
indicators, only those who rented accommodation from a local authority remained
at a 50% increased risk of LUADT cancer even after adjustment for all the
behaviours (RR 1.50 95% CI 1.05, 2.16). This study demonstrated that smoking is a
major inequality issue and a significant cancer risk which is socially patterned.

Further analytical research is required to fully understand the pathways and
mechanisms between socioeconomic circumstance and lung and upper aerodigestive
cancer risk. This thesis suggests that when monitoring socioeconomic
inequalities and cancer risk, it is less effective to focus on all cancer as a group
given the mix of diseases resulting from very different aetiological processes, some
associated with high SES and others with low SES. It also suggests that both
individual and area measures of SES are valid measures and are required to capture
the multi-dimensional nature of SES as well as the life-course and
intergenerational implications of SES. In addition to this “multi-dimensional”
attribute to SES, it is essential to consider multiple low social circumstances
occurring simultaneously and therefore compounding vulnerability to cancer risk.
Behaviours, particularly smoking and alcohol, explained much of the elevated lung
and upper aero-digestive tract cancer risk for individual SES indicators. Clearly, in
this context, smoking is a major inequality issue and a significant cancer risk.

This thesis provides useful insights for raising the issue of inequalities in cancer,
for advocacy and for building policy and interventions to tackle inequalities in
cancer incidence. Policies need to focus on more broadly upstream causes.
Traditionally, these policies have been focused on downstream behaviours (e.g.
public space smoking ban and alcohol minimum pricing), but upstream policies that
take on the fundamental political decisions regarding the distribution of income,
wealth and power are required at both Westminster and Holyrood and beyond.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: socioeconomic inequalities, poverty, deprivation, cancer incidence, lung, head and neck, oesophageal, tumour, malignancy, neoplasm, social determinant.
• Subjects: H Social Sciences > HV Social pathology. Social and public welfare; R Medicine > RA Public aspects of medicine > RA0421 Public health. Hygiene. Preventive Medicine; R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer)
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Medicine, Dentistry & Nursing
• Supervisor's Name: Conway, Professor David I. and McMahon, Dr. Alex
• Date of Award: 2018
• Depositing User: Dr Katharine H Sharpe
• Unique ID: glathesis:2018-8887
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 12 Mar 2018 10:56
• Last Modified: 12 Mar 2018 11:13
• URI: https://theses.gla.ac.uk/id/eprint/8887