|Year : 2019 | Volume
| Issue : 1 | Page : 33-38
Trends and future burden of tobacco-related cancers incidence in Delhi urban areas: 1988–2012
Rajeev Kumar Malhotra1, Nalliah Manoharan2, Omana Nair3, S V S Deo4, Gourva Kishore Rath5
1 Scientist-II, Delhi Cancer Registry, Dr. BRAIRCH, AIIMS, New Delhi, India
2 Scientist-IV, Delhi Cancer Registry, Dr. BRAIRCH, AIIMS, New Delhi, India
3 Scientist-IV, Department of Radiotherapy, AIIMS, New Delhi, India
4 Professor and Head, Department of Surgical Oncology, Dr. BRAIRCH, AIIMS, New Delhi, India
5 Professor and Chief, Department of Radiotherapy, Dr. BRAIRCH, AIIMS, New Delhi, India
|Date of Web Publication||12-Mar-2019|
Dr. Rajeev Kumar Malhotra
Room Number 24, Delhi Cancer Registry, Dr. BRAIRCH, AIIMS, New Delhi
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Introduction: Tobacco products are the major contributors for various cancers and other diseases. In India, tobacco-related cancers (TRCs) contribute nearly half of the total cancers in males and one-fifth in females. Objective: The objective of the study is to investigate 25-year trends and projection of TRCs for 2018–2022. Methods: Joinpoint analysis was performed to assess the trends of TRCs on world age-adjusted rates. Age-period-cohort model with power link function was performed to project the future incidence burden of TRCs in urban Delhi. Results: During the 25 years, a total of 67,129 TRCs (53,125 males and 14,004 females) were registered which was 25.4% of total cancer cases registered. Males contributed 39.1% and females 10.8% of total cases. In males, TRCs declined significantly from 1988 to 2003 with estimated annual percentage change (EAPC) = −0.91% and thereafter increasing trend was observed with EAPC = 3.42%, while in females, the EAPC values were 2.2% and 3.54% respectively for the same period. The total burden of TRCs will be doubled in 2018–2022 with around 46% change due to cancer risk and around 54% due to population age and size in both the genders. The average annual count in males will be 7310 in 2018–2022 as compared to 3571 in 2008–2012 while in females this count will be increased to 2066 from 955 based on recent slope. Conclusion: The incidence of TRCs is increasing due to increase in population age, size, and factors other than population. TRCs are the preventable cancers, and load of these cancers can be controlled with strictly adhering the policy and acts.
Keywords: Age-period-cohort model, joinpoint, projection, tobacco cancer, trend
|How to cite this article:|
Malhotra RK, Manoharan N, Nair O, Deo S V, Rath GK. Trends and future burden of tobacco-related cancers incidence in Delhi urban areas: 1988–2012. Indian J Public Health 2019;63:33-8
|How to cite this URL:|
Malhotra RK, Manoharan N, Nair O, Deo S V, Rath GK. Trends and future burden of tobacco-related cancers incidence in Delhi urban areas: 1988–2012. Indian J Public Health [serial online] 2019 [cited 2020 Aug 3];63:33-8. Available from: http://www.ijph.in/text.asp?2019/63/1/33/253896
| Introduction|| |
Tobacco is a major preventable risk factor not only for various cancers but also for other diseases such as hypertension and cardiovascular diseases. The National Cancer Registry Programme, India, defined 10 tobacco-related cancer (TRC) sites which are lip (C00), tongue (C01–C02), mouth (C03–C06), oropharynx (C10), hypopharynx (C12–C13), pharynx (C14), esophagus (C15), larynx (C32), lung (c33–C34), and urinary bladder (C67). The tobacco product almost contained 28 chemicals carcinogenic constituents. In India, tobacco added a major burden on the total cancer cases and nearly 45% of male's cancer and 20% of female's cancer are due to tobacco use. The proportion of TRCs varies in India according to the geographical region. Meghalaya state has the highest proportion of TRCs for both genders, 65.2% in males and 42.3% in females while Naharlagun has lowest TRCs proportion in males (24.4%) and Pasighat as lowest in females (6.9%). The Delhi urban area has TRCs proportion around 40.7% in men and 11.9% in women which is ranked 14th and 24th among the 27th population-based cancer registry. TRCs are preventable cancers and burden of TRCs can be reduce, but in India, it is a big challenge because most of the stakeholders are illiterate population. The aim of the present study is to study the temporal trend of TRCs incidence in Delhi urban from 1988 to 2012 period and projection of future burden of TRCs during 2018–2022. The assessment of future burden of TRCs will help (1) to plan the future service and allocation of resources to combat the TRCs burden and (2) to help the policymakers to develop the strategies to prevent and control tobacco products. Various statistical methods have been proposed for projection of cancer incidence and mortality in various countries.,,, We applied the method proposed by Moller et al. which considered age-period-cohort (APC) model with power link function instead of the logarithmic link function.,
| Materials and Methods|| |
The study is based on data collected by Delhi PBCR, one of the oldest cancer registries of India. The data on TRCs incidence during 1988 through 2012 are obtained from this registry and does not need ethical requirement for publication of this research paper. The latest available data in Delhi PBCR are for the year 2012. Date were summarized according to gender into five 5-year periods (1988–1992, 1993–1997, 1998–2002, 2003–2007, 2008–2012) and 16 age groups of 5-year interval (0–4, 5–9, 10–14, 15–19,..., 75+), and 20 birth cohort of 5-year interval. The 5-year interval was selected because NORDPRED software is based on a step function of a 5-year interval. This registry fulfilled the IARC data quality standards, and the data were published in Cancer Incidence in Five Continents volume IX and volume X., The international classification of disease code 9th revision was used for the time period 1988–2000 and 10th revision was utilized for 2001–2012. According to 2011 census, the total population of Delhi was 1,67,53,235 with 97.5% of people living in the urban area. The population at risk during of respective years and population for 2018–2022 was estimated from the 1981, 1991, 2001, 2011 census reports of India using different distribution method.
Age-adjusted incidence rates (AARs) per 100,000 was determined by the direct method using the world standard population as reference. The cumulative risk (0–74) and life term risk was computed for both male and female. Life term risk expressed as one out of the total number of persons who develop cancer any time during the whole lifespan was determined on 25-year data.
Joinpoint regression model using joinpoint regression program (Version 220.127.116.11-January 4, 2017, NIH, Division of cancer controls and population sciences, USA) was applied to study the time trends on the AARs according to gender. The trend coefficient was considered statistically significant increase or decrease if P < 0.05.
APC model with power link recommended by the Moller et al. was applied for projection using the package NORDPRED  available in statistical computing R-software version 3.3.3 (R-core Team, Vienna, Austria). This method produces more accurate prediction than logarithmic because latter reflect the exponential growth and overestimate the count over the time. The power link function fortified against this problem and widely applied model in cancer projection worldwide.,,, The NORDPRED software (Cancer Registry of Norway, Norway) was used to predict the incidence and mortality. The (APC) power link model can be written as follows:
Where Rap is the incidence rate in age group a and period p. Aa is the parameter attributable to the age group a, D is the common drift parameter evaluating the linear part of the trend in period and cohort, Pp is the nonlinear effect associated with period p and Cc is the nonlinear effect attributable to cohort c.
The lower age limit for this model was considered where at least 25 subjects were available in all five 5-year periods to get the reliable estimate of projection. The age groups below this number will be projected using the past 10-year average rates. The model for observed rates for TRCs was fitted by applying the four and five 5-year periods depending on the goodness of fit test by Chi-square test and also comparing the Akaike information criteria. Thus, lower age chosen for the model for males was 15–19 years and for females was 20–24 years. The Chi-square goodness of fit was applied between model fitted and observed value for each 5-year period in 16 age groups.
The average trends might provide inaccurate projection when there was a significant sharp change in the AARs trend. The decision on selection of average and recent trend of slope for drift was taken after comparing the above model with the following model:
The significant coefficient of S indicates a departure from the linearity and projection will be estimated on the basis of a recent 10-year trend instead of average trend to project the drift component D.
The percentage change in cancer incidence counts during the projection years was divided into two components, change in cancer incidence due to other than population and changes due to age and size of the population, determined according to the method given by Moller et al. The age-standardized incidence rates were calculated using the world population weights.
| Results|| |
During the 25 years, a total of 67,129 TRCs (53,125 males and 14,004 females) were registered which accounts for 25.4% of total cancer cases registered. The TRCs contributed 39.1% and 10.8% of total cases in males and females, respectively. Males dominated in TRCs than females with ratio of 3.8-1. The median age of diagnosis of TRCs in urban Delhi during the 25-year period was 59.3 years and 58.4 years for males and females, respectively. Lung cancer contributed one-fourth of the TRCs cases in males followed by larynx (15.9%) and tongue (14.7%). In females, the lung cancer share was 23% followed by esophagus (22.7%) and mouth (14.8%). In males, AARs increased by 39% from 46.14 per 100,000 in 1988 to 64.19 per 100,000 in 2012. However, in females, AARs increase was marginal from 17.08–18.39 during this period.
The age-specific incidence of TRCs showed an exponential increase in males from 35 to 75 years and while a linear trend was observed in females from 35 to 65 years and plateau after that [Figure 1]. One in 16 among males and 1 in 55 among females were likely to develop cancer-related to tobacco in the absence of any competing risk.
|Figure 1: Age-specific incidence rate of tobacco-related cancers by gender, 1988–2012.|
Click here to view
One joinpoint was observed in both males and females [Figure 2]. The estimated annual percentage change (EAPC) revealed a declining trend during 1988–2002 with −0.91% (95% CI: −1.63 to −0.18; P = 0.02) in males and −2.2% (95% CI: −3.11 to −1.27; P < 0.001) in females respectively. In the recent decade (2003–2012), a rising significant trend was observed in AARs in both males and females with EAPC 3.42% (95% CI 2.3–4.56; P < 0.001) and 3.54% (95% CI: 2.09–5.00; P < 0.001), respectively. The proportion of TRCs to total cancer also showed a significant increasing linear trend in males from 35.09% in 1988 to 41.67% in 2012. No significant trend was observed in females albeit marginally increased in TRCs percentage was found from 10.9% to 11.9%.
|Figure 2: The temporal trend of age-adjusted incidence rate of tobacco-related cancers, 1988–2012.|
Click here to view
In males, four 5-year periods showed a better fit while in females all five 5-year periods performed better in model fitting. The model with a recent trend for drift revealed significantly different with average model represents that former model provides a better estimate for projection than latter. The linear drift value modified was as D and 0.75 D for projecting the next two 5-year periods.
[Table 1] represents the actual counts, crude rate, and AARs for each 5-year period for observed and projected period in Delhi urban by gender. [Figure 3] represents trend (1988–2012) and projection (2013–2022) for age-specific cancer incidence rates for total TRCs according to gender. The solid line represented observed rates and dotted line showed projected rates.
|Table 1: Recorded and predicted tobacco-related cancers incidence cases by gender in Delhi Urban|
Click here to view
The predicted male population in Delhi urban will increase by about 23% up to 2018–2022 from 2008 to 2012 and predicted TRCs burden in 2018–2022 showed an increase of more than 100% in TRCs cancer as compared to 2008–2012. The average annual incidence count in males will be 7,310 in 2018–2022 from 3,571 in 2008–2012 while in female this will be 2,066 from 955. The increase among the males amounts to 56.67% due to the impact of demographic changes and an increase of 47.98% due to the risk factors. Among the females, the population will increase by 30% and overall TRCs count will be 116% in 2008–2012 when compared to 2008–2012, out of this, 62% due to demographical changes and 54% due to change in cancer risk due other than demographical changes [Table 1].
| Discussion|| |
This study provides the temporal trend of incidence of TRCs and projection of TRCs in Delhi urban area. The present model incorporates the non-linear variation components of period and cohort in the projection of TRCs, and this method was also applied in predicting the future burden in Tamil Nadu on the top ten cancer sites. The future burden estimates help the health service providers to update the infrastructure and allocation of resources to combat the cancer control.
Results revealed that the contribution of TRCs to total cancer increased during the 25 years in males while marginally raised in females. The latest decade showed an upward rise in trend in both genders. Our model revealed that recent trend slope provides significantly better projection than average slope. The average annual count of newly diagnosed cases using recent slope will be double in both males and females on 2018–2022 as compared to 2008–2012. The AARs of TRCs are projected to change from 2008–2012 to 2018–2022. The rates are expected to increase by 29% in males from 59.08–76.20 per 100,000, and to increase by 23% from 16.82–21.92 in females [Table 1].
In our study, the increase in predicted count could be attributed approximately, equally due to changes in demographic effect and changes due to risk. However, the Chennai study showed more change to risk rather than demographic effect where Manitoba study showed change due to demographic effect is higher than changes due to risk.,
According to GLOBOCAN report of 2012, the developed region had more incidence of TRCs as compared to the developing countries. Globally, 32% of total cancers were attributable to TRCs in males while in female this percentage was 15%. Delhi contributed a higher percentage in males (41%) and lower percentage (12%) in females according to 2012 data. In males, world TRCs AARs was 64.9 per 100,000, with one in 14 individual will developed TRCs while in India AARs 40 per 100,000 and in Delhi AARs 57 per 100, 000 in the 2012. The highest AARs of TRCs among males was found in 147 per 100,000 in Hungry. In females, world AARs 22.6 per 100,000 while India AARs 12.6 per 100,000 and in Delhi 18.3 per 100,000.
India is the third largest producer and second largest consumer of tobacco products in the world. There are various varieties of tobacco products and prevalence of these products also varies in India due to regional and cultural variations. In some parts, the prevalence of smoking is more while in others smokeless tobacco dominates. In female usages of smokeless, tobacco is more than six times than smokers and the rural area are more prone than urban. The incidence of TRCs also varies widely as per geographical location and gender in India. About 24%–65% of total cancers among the males and 7%–42% among the females are related to tobacco. As per the recent report of the Global Adult Tobacco Survey 2016–2017, 28.6% of all adult (15+) either smoke tobacco and/or use smokeless tobacco, out of these intake of smokeless tobacco is 21.4% as compared to smoking 10.7%. Although overall results of GATS-2 showed a decline by absolute 6% from 34.6% (GATS-1) to 28.6% (GATS-2) the impact of this decline will not affect the future burden of TRCs up to 2022.
The comparison of another two previous survey Special Fertility and Mortality (1998) and GATS-1 (2010) reflect a marginal decrease in overall prevalence of smoking among the men aged 15–69 years in urban population (21% vs. 20.4%) while in Delhi urban there was a significant increase in the age-standardized prevalence from 13.9% in 1998 to 31.4% in 2010. Among females, ASP of any smoking in aged 15–69 years rose from 1.4% in 1998 to 2.7% in 2010.
Smoking is the strongest risk factor of lung cancer and increased the risk of lung cancer in male by 9–10 fold as compared to nonsmokers  and contributing etiological agent for remaining TRCs. Lung cancer in never smoker is also increasing, however not much studies were available on the nonsmoker, but now the risk of other factors apart from smoking is area of concern and need to be investigated., Better economic growth especially in urban area led sedentary lifestyle and intake of high calorific food may be some factors associated with lung cancer. The other nontobacco factors for lung cancer may be indoor and outdoor pollution, exposure to radon and asbestos.
The reduction in smoking has been cited as the single most important factor for reducing the AARs trend in England. In India, although the usage of smoking and smokeless tobacco showing a decline in the pattern, the effect of decline prevalence of tobacco product will reflect in the next 15 years. The usage is still very high and needs to be addressed by devising the appropriate strategy to spread the awareness programs by means of public education campaigns, motivation for tobacco cessation especially in young women and men because these are more vulnerable population.
The lip, mouth, and tongue cancer together comes closer to lung cancer cases. In Delhi, Mumbai, Kolkata, and Chennai after lung cancer, mouth and tongue are the most common cancer related to TRCs. The smokeless tobacco is major cause of more than 60% of cases of the oral cavity, and in India, more than 30 different varieties of smokeless tobacco are available. As per GATS, one another major challenge is areca nut chewing along with betel leaves and common usage of gutka and khaini or lime mixture with tobacco.
Esophagus contributed about 13% of TRCs in Delhi urban during 25 years. This is the only TRCs which has showed the ratio of 1–1.76 between males and females. The study showed apart for tobacco other risk factors for this cancer are foods rich in nitrogenous components and less intake of fruit and vegetable.
Population-based cancer registry has a higher quality of data and covers 97.5% of Delhi population and data were also included in IARC publication., Long-term data were utilized for evaluating the trends and estimating the future burden.
The NORDPRED method developed in the Nordic countries recommended for cancer projection but have some limitations. The projection rates are based on the assumption of continuity of the past trend in future. The projection is more accurate for a shorter period than the longer periods. Our study projections are for short-term, as trend more likely to change over the long periods, so chance of error in projection is less in our study. Another important factor that may influence the projection is an accurate forecast of the population. In the study, the population for the next 10 years were projected on the basis of growth between 2001–2011 using method described by Takiar and Shobana which considered the growth rate over 10-year but not considered immigration or fertility change, due to this our projection may be over or underestimate than actual.
| Conclusion|| |
The projections of cancer incidence, although inherently subjective and may not be well matched with future actual, but can provide a baseline for the future planning. The methodology used in this article to estimate the counts minimize the subjectivity. TRCs are the preventable cancers, and load of these cancers can be controlled with strictly adhering the policy and Acts developed by the authorities. The young population that is more vulnerable to the smoking and smokeless products need to be properly educated through the awareness programs. Our projection estimates provide a baseline platform for future cancer control planners to allocate the resources. Public health policymakers and Delhi health professionals can assess the future burden of TRCs in Delhi urban and develop the strategy to combat this deadly preventable disease.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Banerjee SC, Ostroff JS, Bari S, D'Agostino TA, Khera M, Acharya S, et al.
Gutka and Tambaku Paan use among South Asian immigrants: A focus group study. J Immigr Minor Health 2014;16:531-9.
National Cancer Registry Programme. Three Year Report of Population based Cancer Registries: 2012-2014 Report of 27 PBCRs in India. Bangalore, India: National Cancer Registry Programme; 2016. p. 9-16.
Kaur J, Jain DC. Tobacco control policies in India: Implementation and challenges. Indian J Public Health 2011;55:220-7.
] [Full text]
Møller B, Fekjaer H, Hakulinen T, Tryggvadóttir L, Storm HH, Talbäck M, et al.
Prediction of cancer incidence in the Nordic countries up to the year 2020. Eur J Cancer Prev 2002;11 Suppl 1:S1-96.
Moller B, Fekjaer H, Hakilinen T, Tryggvadottir L, Strom HH, Talback M, et al
. Prediction of cancer incidence in the Nordic countries: Empirical comparison of different approaches. Stat Med 2003;22:2751-66.
Clements MS, Armstrong BK, Moolgavkar SH. Lung cancer rate predictions using generalized additive models. Biostatistics 2005;6:576-89.
Katanoda K, Kamo K, Saika K, Matsuda T, Shibata A, Matsuda A, et al.
Short-term projection of cancer incidence in Japan using an age-period interaction model with spline smoothing. Jpn J Clin Oncol 2014;44:36-41.
Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, et al
. Cancer Incidence in Five Continents. IARC Scientific Publication no. 164. Vol. 9. Lyon, France: IARC; 2007. p. 229.
Forman D, Bray F, Brewster DH, Mbalawa CG, Kohler B, Pineros M, et al
. Cancer Incidence in Five Continents. IARC Scientific Publication no. 164, Vol. 10. Lyon, France: IARC; 2014. p. 550.
Takiar R, Shobana B. Cancer incidence rates and the problem of denominators – A new approach in Indian cancer registries. Asian Pac J Cancer Prev 2009;10:123-6.
Esteve J, Benhamou E, Raymond L, editors. Statistical Methods in Cancer Research: Descriptive Epidemiology. Scientific Publications No 128. Vol. 4. France: IARC; 1994. p. 313.
Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med 2000;19:335-51.
Møller H, Fairley L, Coupland V, Okello C, Green M, Forman D, et al.
The future burden of cancer in England: Incidence and numbers of new patients in 2020. Br J Cancer 2007;96:1484-8.
Lee TC, Dean CB, Semeneiw R. Short term cancer mortality projection: A comparative study of prediction methods. Stat Med 2011;30:3387-402.
Swaminathan R, Shanta V, Ferlay J, Balasubramanian S, Bray F, Sankaranarayanan R, et al.
Trends in cancer incidence in Chennai city (1982-2006) and statewide predictions of future burden in Tamil Nadu (2007-16). Natl Med J India 2011;24:72-7.
Nowatzki J, Moller B, Demers A. Projection of future cancer incidence and new cancer cases in Manitoba, 2006-2025. Chronic Dis Can 2011;31:71-8.
Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, et al
. GLOBOCAN 2012 v1.1, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No 11. Lyon, France: International Agency for Research on Cancer; 2014. Available from: http://www.globocan.iarc.fr
. [Last accessed on 2017 Aug 31].
International Institute for Population Sciences, Ministry of Health and Family Welfare, Government of India. Global Adult Tobacco Survey India Report (GATS India), 2009–10. New Delhi, Mumbai: MoHFW, Government of India, IIPS; 2010.
Registrar General of India. Special Fertility and Mortality Survey 1998: A Report on. 1 Million Homes. New Delhi, India: Registrar General of India; 2005.
Mishra S, Joseph RA, Gupta PC, Pezzack B, Ram F, Sinha DN, et al.
Trends in bidi and cigarette smoking in India from 1998 to 2015, by age, gender and education. BMJ Glob Health 2016;1:e000005.
Dela Cruz CS, Tanoue LT, Matthay RA. Lung cancer: Epidemiology, etiology, and prevention. Clin Chest Med 2011;32:605-44.
Sasco AJ, Secretan MB, Straif K. Tobacco smoking and cancer: A brief review of recent epidemiological evidence. Lung Cancer 2004;45 Suppl 2:S3-9.
Krishnamurthy A, Vijayalakshmi R, Gadigi V, Ranganathan R, Sagar TG. The relevance of “Nonsmoking-associated lung cancer” in India: A single-centre experience. Indian J Cancer 2012;49:82-8.
] [Full text]
Noronha V, Dikshit R, Raut N, Joshi A, Pramesh CS, George K, et al.
Epidemiology of lung cancer in India: Focus on the differences between non-smokers and smokers: A single-centre experience. Indian J Cancer 2012;49:74-81.
] [Full text]
Mistry M, Parkin DM, Ahmad AS, Sasieni P. Cancer incidence in the United Kingdom: Projections to the year 2030. Br J Cancer 2011;105:1795-803.
Domper Arnal MJ, Ferrández Arenas Á, Lanas Arbeloa Á. Esophageal cancer: Risk factors, screening and endoscopic treatment in Western and Eastern countries. World J Gastroenterol 2015;21:7933-43.
[Figure 1], [Figure 2], [Figure 3]