Sage Journals: Discover world-class research

Abstract

Objectives

Obesity is associated with an increased cancer incidence and mortality and therefore cancer screening is particularly important for obese individuals. However, some US studies find lower screening uptake in this group. This study explored whether rates of breast and colorectal screening in England are lower for obese than healthy weight individuals.

Setting

Data were from the English Longitudinal Study of Ageing (ELSA).

Methods

We analysed data from adults who were eligible to have been invited to the UK national screening programmes for breast or colorectal cancer (CRC) in the last five years, and had been given the screening module in Wave 5 of ELSA (N = 1804 for CRC screening, N = 2401 for breast cancer screening). Weight and height were measured by a nurse at Wave 4 (two years earlier). Logistic regression was used to calculate the odds of breast and CRC screening (ever) for participants in higher weight categories (Body Mass Index [BMI] ≥ 25) compared with healthy weight individuals (BMI <25), controlling for socio-demographic variables.

Results

Of ELSA participants, 63% reported CRC screening, and 92% of the women reported breast cancer screening. Obesity was associated with lower CRC screening, but effects were strongest for class III obesity (BMI ≥40) (45% screened; OR = 0.48, 95% CI = 0.32–0.93, P = .029). There was no association between weight status and breast cancer screening.

Conclusion

Severe obesity appears to be a deterrent to CRC screening but not breast cancer screening. Targeted interventions may be required to promote CRC screening uptake in this group, which already has a heightened risk as a consequence of weight.

Keywords

cancer screening obesity breast cancer colorectal cancer body mass index

Introduction

Obesity is associated with increased incidence and mortality of certain cancers.^1,2 The biological mechanisms linking obesity and cancer are still being uncovered, although sex hormones and growth factors are both known to play a role.^3–5 Differences in early detection behaviours - screening and symptomatic presentation - could play a part in raised mortality.

Two recent meta-analyses of US studies have explored relationships between obesity and mammography attendance⁶ and colorectal cancer (CRC) screening.⁷ For mammography, there was an inverse relationship overall between obesity and recent mammography, and women with class III obesity were significantly less likely to have received mammography than normal weight women. For CRC screening, while a higher Body Mass Index (BMI) was not associated with lower rates of screening overall, white women with class I, II, or III obesity were less likely to have been screened than normal weight women, and white men with class II obesity were also less likely to have been screened compared with those with a normal BMI. However, none of the studies included in the review used objectively measured weights, which may have had the effect of ensuring that the groups identified as obese were even heavier than they appeared.

With the exception of three studies - one study in Canada⁸, one in Korea⁹, and one in Switzerland¹⁰- studies in this area have all been based on US samples. It is important to determine if weight status differences are found in a country where organized national screening programmes are offered for both breast and CRC. The aim of the present study was therefore to examine differences in breast and CRC screening behaviour in England in a sample of age eligible adults with objectively measured weight status. We hypothesized that screening would be lower for obese individuals than for those in the healthy weight range.

Methods

Study population

The data for these analyses are from the subset of participants who were asked questions about cancer screening in Wave 5 of the English Longitudinal Study of Ageing (ELSA), carried out in 2010/2011. ELSA is a cohort of individuals aged 50 and over, living in England, who have been followed up every two years since 2002, with weight and height measurements taken by nurses in alternate waves (Waves 0, 2 & 4). The total ELSA sample in Wave 5 consisted of 13,428 participants.^11,12 The screening module was introduced part way through Wave 5. Valid nurse-measured heights and weights from Wave 4 (2008/2009) were available for 77.5% of Wave 5 participants. The current study was exempt from ethical approval under the guidelines of the University College London Research Ethics Committee because it was secondary analysis of ELSA data that was provided to researchers in an anonymized format, with no identifiable information included. Data from the 2 waves are linked using unique individual serial numbers.

Measures

Anthropometric, demographic and health measures

Height to the nearest millimetre was measured using a portable stadiometer, and weight was calculated to the nearest 0.1kg with Tanita THD-305 portable electronic scales. These values were used to calculate BMI using the standard formula, and BMI was categorized into normal weight (BMI< 25), overweight (BMI 25–29.9), class I obesity (BMI 30–34.9), class II obesity (BMI 35–39.9) and class III obesity (BMI ≥ 40).

Age, sex, ethnicity, marital status, and non-pension wealth (a sensitive indicator of socioeconomic status (SES) in this age group) were reported. For our analyses, age was categorised as <65 vs. ≥65, marital status as married vs. unmarried (single/never married, separated, divorced, or widowed), ethnicity as white vs. non-white, and wealth was split into quintiles.

Participants also reported any diagnosis of cancer, angina, diabetes, stroke, coronary heart disease, or other limiting long standing illness.

Cancer screening

As part of the UK National Health Service (NHS) breast screening programme, women aged 50 to 70 (extended to 74 in some areas) are invited for screening by mammography every three years. The NHS Bowel Cancer Screening Programme was fully implemented in England in 2010. Men and women aged 60–69 (being extended to 74) are sent a Faecal Occult Blood Testing (FOBT) kit every two years to complete at home and return for analysis.

The screening questions were prefaced with: ‘I would like to ask you about any screening you may have had to check for different types of cancer’. Women were asked: Have you ever had a mammogram (x-ray of your breasts)’ and (if yes) ‘How long ago was your most recent mammogram’; followed by ‘What was the reason for your most recent mammogram’, which included the response option ‘Routine screening by the NHS Breast Screening Programme’. All participants were asked ‘Have you ever completed a home testing kit for screening bowel cancer’, and (if yes) ‘How long ago was your most recent test’; followed by ‘Was this test part of the NHS Bowel Cancer Screening Programme’.

We were specifically interested in those who reported NHS screening, as other offers of screening may have been made because of a specific health concern (e.g. a non-routine referral or follow-up cancer care) and it is likely that individuals would respond differently to these. Therefore, for the purposes of our analyses, only NHS screening was classified as screening (excluding 23 potential participants from the CRC screening analysis, and 80 from the breast cancer screening analysis). The present analyses used ‘ever been screened’ as the outcome.

Statistical analysis

We restricted our age range to those who would have been invited to NHS screening within the last five years, taking account of age eligibility and (for CRC screening) the date that the programme was fully implemented across the country. For CRC screening, this was age 62–69, and for breast cancer screening 50–72. Multivariable logistic regressions including age, sex (CRC screening only), ethnicity, marital status, and SES (indicated by non-pension wealth) were used to calculate the odds of CRC and breast cancer screening for individuals in each higher weight group (overweight, class I obesity, class II obesity, class III obesity) compared with those with BMI <25. Polynomial contrasts were used to examine overall trends.¹³ Because co-morbidities may influence screening uptake^4,14–17, analyses were repeated after controlling for a diagnosis of cancer, angina, diabetes, stroke, coronary heart disease, or other limiting long standing illness. To explore potential gender differences, we checked for an interaction between weight status and sex for CRC screening. Analyses were carried out using IBM SPSS Statistics version 19, with a two-sided P value <0.05 determining statistical significance.

Results

CRC Screening

Respondents for the analysis of CRC screening were predominantly married (74%) and white (98%). The mean (standard deviation [SD]) age was 65.0 (2.2) and the mean (SD) BMI was 28.2 (5.1). Overall, 63% of this sample reported ever having had CRC screening.

The results of the multivariable logistic regression model, controlling for demographic characteristics, are shown in Table 1. Respondents with class III obesity were significantly less likely to have been screened than those with a BMI < 25 (45% vs. 66%; OR = 0.48, 95% CI = 0.25–0.92, P = .028), and there was a trend towards lower screening rates among class II obese (58% vs. 66%; OR = 0.79, 95% CI = 0.49–1.29, P = .342), but being overweight or having class I obesity did not predict CRC screening (67% and 66% screened respectively). Polynomial contrasts revealed a significant linear trend for BMI on screening (P = .011); as weight status increased, screening rates decreased. There was also a significant quadratic trend (P = .044), which reflects that, as weight status increased, the difference in screening rates increased. There was no interaction between BMI and sex for CRC screening (P = .837).

Table 1.

Multivariable logistic regression models predicting screening.

	CRC Screening (n = 1373)				Breast Cancer Screening (n = 1618)
	Total N (% screened)	OR	CI	P	Total N (% screened)	OR	CI	P
Sex
Male	620 (61)	1.00	–	–	–	–	–	–
Female	753 (69)	1.53	1.21–1.93	0.001	1618 (94)	–	–	–
Age
<65	634 (66)	1.00	–	–	1045 (95)	1.00	–	–
≥65	739 (65)	0.97	0.77–1.22	0.77	573 (93)	0.76	0.50–1.17	0.21
Wealth quintile
1	190 (48)	1.00	–	–	240 (90)	1.00	–	–
2	225 (66)	2.01	1.35–3.04	0.001	339 (94)	1.70	0.90–3.21	0.10
3	281 (72)	2.54	1.71–3.79	<0.001	321 (93)	1.34	0.72–2.51	0.36
4	319 (69)	2.20	1.51–3.28	<0.001	342 (95)	1.67	0.86–3.26	0.13
5	358 (67)	2.00	1.36–2.92	<0.001	375 (97)	2.59	1.23–5.43	0.01
Marital status
Married	1009 (67)	1.00	–	–	1107 (95)	1.00	–	–
Unmarried	364 (60)	0.82	0.63–1.08	0.16	511 (91)	0.60	0.38–0.94	0.02
Ethnicity
White	1346 (66)	1.00	–	–	1575 (94)	1.00	–	–
Non-white	27 (44)	0.38	0.17–0.84	0.02	43 (95)	1.25	0.29–5.43	0.77
BMI
<25	371 (66)	1.00	–	–	525 (93)	1.00	–	–
25–29.9	594 (67)	1.11	0.84–1.48	0.46	561 (95)	1.50	0.90–2.50	0.12
30–34.9	277 (67)	1.10	0.79–1.54	0.59	330 (95)	1.49	0.82–2.71	0.19
35–39.9	89 (58)	0.79	0.49–1.29	0.34	137 (93)	1.23	0.57–2.65	0.60
≥40	42 (45)	0.48	0.25–0.92	0.03	65 (94)	1.38	0.47–4.07	0.55

OR = odds ratio; CI = 95% confidence interval; Percentages shown are adjusted for age, sex (CRC screening only), ethnicity, non-pension wealth and marital status.

Consistent with other studies^18–20, women were more likely than men to have been screened (69% vs. 61%; OR = 1.53, 95% CI = 1.21–1.93; P < .001), those in the higher quintiles for wealth were more likely to have been screened than the bottom quintile (66–72% vs. 48%; all P values ≤ .001), and non-white respondents were less likely to have been screened than white respondents (66% vs. 44%; OR = 0.38, 95% CI = 0.17 to 0.84, P = .017).

We repeated the analyses controlling for a diagnosis of cancer, angina, diabetes, stroke, coronary heart disease, or other limiting long standing illness, and results were essentially unchanged.

Breast Cancer Screening

Women in the sample analyzed for breast cancer screening were predominantly married (70%) and white (96%). Their mean (SD) age was 61.45 (5.54) and mean (SD) BMI was 28.26 (5.81). Over 90% reported ever having had breast cancer screening.

The results of the multivariable logistic regression model are shown in Table 1. Rates were broadly similar across categories, and showed no evidence of a drop-off at higher weight levels (normal weight: 93%, class III obesity: 94%). Consistent with previous research¹⁸, women in the highest wealth quintile were more likely to have been screened than those in the bottom quintile (97% vs. 90%; OR = 2.59, 95% CI = 1.23–5.43, P = .012). Unmarried participants were significantly less likely to have been screened than those who were married (91% vs. 95%; OR = 0.60, 95% CI = 0.38–0.94, P = .024).

We repeated the analyses controlling for a diagnosis of cancer, angina, diabetes, stroke, coronary heart disease, or other limiting long standing illness. As with CRC screening, results were essentially unchanged.

Discussion

In this cohort of adults in England, overall rates for ever having had breast or CRC screening were 92% and 63% respectively, comparable with other population-based survey data.²¹ The lower rates of CRC screening compared with breast cancer screening also corresponds with differences in screening uptake rates for the two types of cancer in Britain.^18,22 In line with our hypothesis that screening rates would be lower for obese individuals, respondents with class II/III obesity were less likely than normal-weight respondents to have had CRC screening, independent of potential demographic and health confounders. There was also a significant linear trend for BMI on CRC screening, with rates decreasing as weight status increased. We found no deterrent effect of obesity on breast cancer screening.

For CRC screening, we also checked for an interaction between BMI and sex on screening rates, but found no differential effect. Previous studies have also explored differences by race^16,23–25, however, we were unable to do so in our sample due to the small numbers of non-white responders (<5%).

A number of explanations have been offered for the lower screening rates among the severely obese. Patients with comorbidities may feel that screening is not a priority²⁶, particularly given that many obese individuals are not aware of their increased cancer risk^27,28, although controlling for co-morbidities did not alter the results in our sample. Obese people often have poor body image, which could cause embarrassment or unease when dealing with health care practitioners^29,30, although FOBT tests are done in the home which should avoid this problem. Individuals with higher BMIs may simply be less engaged in health-seeking behaviours, with CRC screening representing an aspect of this.²⁶ However this does not fit with the fact that, in our sample, obese women were just as likely to have been screened for breast cancer as women with a normal BMI. Lastly, it may be an issue of mobility or flexibility, given the actions required to collect the necessary sample for FOBT. Obesity is associated with mobility problems, particular in older adults^31,32, which could be why we saw an effect of BMI on CRC screening, but not on breast cancer screening, where health professionals are available to help with any mobility problems.

Future research should explore barriers to CRC screening for obese individuals. A better understanding of these issues could help to inform interventions to promote screening equitably, regardless of body size. For example, if mobility or flexibility is a barrier, additional assistance could be provided, or flexible sigmoidoscopy offered, as an alternative method of screening, which may be easier for this population. With the introduction of flexible sigmoidoscopy to the NHS screening programme, it will be important to explore if the differences generalize to this form of CRC screening, particularly as US research has shown differences in the relationship between screening and BMI dependent on the screening modality.²³

This study had a number of limitations. The generally high levels of screening may be a reflection of the ELSA cohort members being conscientious about health matters, and therefore findings cannot be assumed to generalize to the wider UK population. As in other studies, screening was self-reported, so there is a risk that behaviours were over-reported.³³ The percentage screened for CRC represents one (or at most two) round(s) of screening, whereas for breast cancer it may be four or five rounds, and so those who reported not attending breast cancer screening are more likely to be “persistent non-responders” whereas, for CRC, the non-attenders are likely to include both those who would never attend and those who would do so only erratically. Persistent non-attenders may have different characteristics and barriers compared with those who attend only some of the time, which we were unable to explore. ELSA has a lower representation of ethnic minorities than England generally (97% white in ELSA vs. 86% white in the general population³⁴), and because the screening questions were only relevant to a small age range, the ethnic minority numbers were particularly low in our sample. This made it difficult to explore ethnic differences, other than through the comparison of white vs. non-white. An advantage of our study was that it included objective measurements of height and weight, because dependence on self-reported BMI has been cited as a limitation of the research in this area.^6,7

The results of this study highlight the potential impact of weight on participation in CRC but not breast cancer screening, in England. To our knowledge, this is the first investigation of screening uptake by weight status in the English population, and it is therefore important that the findings are replicated. As they stand, they suggest that severely obese individuals may be doubly disadvantaged, with a higher underlying risk of CRC and lower uptake of screening. A better understanding of the barriers to screening for severely obese individuals would help to inform the development of interventions.

Footnotes

Funding

This study was supported by Cancer Research UK [grant number C1418/A7974].

The funding for ELSA is provided by the National Institute of Aging in the United States, and a consortium of UK government departments co-ordinated by the Office for National Statistics. The developers and funders of ELSA and the Archive do not bear any responsibility for the analyses or interpretations presented here.

Acknowledgements

The data were made available through the UK Data Archive (UKDA). ELSA was developed by a team of researchers based at the NatCen Social Research, University College London and the Institute for Fiscal Studies. The data were collected by NatCen Social Research.

References

Renehan

Tyson

Egger

Heller

Zwahlen

. Body-Mass Index and Incidence of Cancer: A Systematic Review and Meta-Analysis of Prospective Observational Studies. Am J Health Promot 2008; 23(2): 153–153.

Calle

Rodriguez

Walker-Thurmond

Thun

. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. New Engl J Med 2003; 348(17): 1625–1638.

Basen-Engquist

Chang

. Obesity and cancer risk: recent review and evidence. Curr Oncol Rep 2011; 13(1): 71–67.

Haakinson

Leeds

Dueck

. The impact of obesity on breast cancer: a retrospective review. Ann Surgical Oncol 2012; 19(9): 3012–3018.

Yang

Wang

. Obesity and risk of colorectal cancer: a systematic review of prospective studies. Plos One 2013; 8(1): e53916–e53916.

Maruthur

Bolen

Brancati

Clark

. Obesity and mammography: a systematic review and meta-analysis. J Gen Intern Med 2009; 24(5): 665–677.

Maruthur

Bolen

Gudzune

Brancati

Clark

. Body mass index and colon cancer screening: a systematic review and meta-analysis. Cancer Epidem Biomar 2012; 21(5): 737–746.

Porter

Inglis

Wood

Veugelers

. Effect of obesity on presentation of breast cancer. Ann Surg Oncol 2006; 13(3): 327–332.

Park

Cho

. Obesity and screening compliance for breast and cervical cancer in Korean women. Asian Pac J of Cancer Prev 2012; 13(7): 3271–3274.

10.

Fischer

Collet

Zeller

Zimmerli

Gaspoz

Giraudon

Rodoni

Cornuz

. Obesity and overweight associated with lower rates of colorectal cancer screening in Switzerland. Eur J Cancer Prev 2013; 22(5): 425–430.

11.

Marmot MB, Banks J, Blundell, R, Lessof C and Nazroo J. Health, wealth and lifestyles of the older population in England: The 2002 English Longitudinal Study Of Ageing. http://www.ifs.org.uk/ELSA (2003, accessed 3 May 2013).

12.

Steptoe

Breeze

Banks

Nazroo

. Cohort Profile: The English Longitudinal Study of Ageing. Int J Epidemiol 2013; 42(6): 1640–1648.

13.

Menard

. Applied Logistic Regression Analysis. Sage University Series Papers on Quantitative Applications in the Social Sciences, 07–106, 2nd ed. Thousand Oaks, CA: Sage, 2002.

14.

Ferrante

Chen

Jacobs

. Breast and cervical cancer screening in obese minority women. J Womens Health 2006; 15(5): 531–541.

15.

Fontaine

Faith

Allison

Cheskin

. Body weight and health care among women in the general population. Arch Fam Med 1998; 7(4): 381–384.

16.

Ostbye

Taylor

Yancy

Krause

. Associations between obesity and receipt of screening mammography, Papanicolaou tests, and influenza vaccination: Results from the health and retirement study (HRS) and the asset and health dynamics among the oldest old (AHEAD) study. Am J Public Health 2005; 95(9): 1623–1630.

17.

Phillips

Smith

Ahn

Ory

Hochhalter

. Correlates of Initiating Colorectal Cancer Screening Beginning at Age 50. J Commun Health 2013; 38(1): 23–30.

18.

von Wagner

Baio

Raine

. Inequalities in participation in an organised national colorectal cancer screening programme: results from the first 2.6 million invitations in England. Int J Epidemiol 2011; 40: 712–718.

19.

Bansal

Bhopal

Steiner

MFC

Brewster

. Major ethnic group differences in breast cancer screening uptake in Scotland are not extinguished by adjustment for indices of geographical residence, area deprivation, long-term illness and education. Brit J Cancer 2012; 106: 1361–1366.

20.

Naylor K, Ward J, Blasé and Polite BN. Interventions to Improve Care Related to Colorectal Cancer Among Racial and Ethnic Minorities: A Systematic Review. J Gen Intern Med 2012; 27(8): 1033–1046.

21.

Hing Lo

Waller

Wardle

von Wagner

. Comparing barriers to colorectal cancer screening with barriers to breast and cervical screening: a population-based survey of screening-age women in Great Britain. J Med Screening 2013; 20(2): 73–79.

22.

NHS Breast screening Programme. Annual Review 2012. http://www.cancerscreening.nhs.uk/breastscreen/publications/nhsbsp-annualreview2012.pdf (2012, accessed 24 May 2013).

23.

Leone

James

Hudson

Campbell

. Obesity predicts differential response to a cancer prevention intervention among African-Americans. Cancer Epidem Biomar 2007; 16(2): 360–360.

24.

Cohen

Murff

Signorello

Blot

. Obesity and colorectal cancer screening among black and white adults. Cancer Cause Control 2012; 23(5): 709–716.

25.

Zhu

Jatoi

Potter

Shriver

. Body mass index and use of mammography screening in the United States. Prev Med 2006; 42(5): 381–385.

26.

Wee

McCarthy

David

Phillips

. Screening for cervical and breast cancer: Is obesity an unrecognized barrier to preventive care? Ann Intern Med 2000; 132(9): 697–704.

27.

Messina

Lane

Anderson

. Body mass index and screening for colorectal cancer: Gender and attitudinal factors. Cancer Epidemiol 2012; 36(4): 400–408.

28.

Cameron

Scully

Herd

Jamsen

Hill

Wakefield

. The Role of Overweight and Obesity in Perceived Risk Factors for Cancer: Implications for Education. J Cancer Educ 2010; 25(4): 506–511.

29.

Olson

Schumaker

Yawn

. Overweight women delay medical care. Arch Fam Med 1994; 3(10): 888–892.

30.

Adams

Smith

Wilbur

Grady

. The relationship of obesity to the frequency of pelvic examinations: do physician and patient attitudes make a difference? Women Health 1993; 20(2): 45–57.

31.

Forhan

Gill

. Obesity, functional mobility and quality of life. Best Pract Res Cl En 2013; 27(2): 129–137.

32.

Vincent

Lamb

. Obesity and mobility disability in the older adult. Obes Rev 2010; 11(8): 568–579.

33.

Jones

Mongin

Lazovich

Church

Yaezel

. Validity of four self-reported colorectal cancer screening modalities in a general population: Differences over time and by intervention assignment. Cancer Epidem Biomar 2008; 17(4): 777–784.

34.

Office of National Statistics. Ethnicity and National Identity in England and Wales 2011, http://www.ons.gov.uk/ons/dcp171776_290558.pdf (2012, accessed 5 June 2013).

Body mass index and cancer screening: Findings from the English Longitudinal Study of Ageing

Abstract

Objectives

Setting

Methods

Results

Conclusion

Keywords

Introduction

Methods

Study population

Measures

Anthropometric, demographic and health measures

Cancer screening

Statistical analysis

Results

CRC Screening

Breast Cancer Screening

Discussion

Footnotes

Funding

Acknowledgements

References