Tobacco smoking behavior and health outcomes among patients with chronic diseases: a comparative study of tobacco-producing and non-producing areas in southern Thailand Authors: Udomsak Saengow, Temesgen Tamirat, Tadesse Lelago Published: 2026 ABSTRACT Tobacco use remains a substantial cause of morbidity among patients with noncommunicable diseases, especially in areas with localized tobacco production. This study looked into the tobacco smoking behavior and, adverse health outcomes, and predictors of smoking behavior among chronic disease patients in both tobacco-producing and non-producing areas of Nakhon Si Thammarat, Thailand. An analytical cross-sectional study was undertaken with 879 patients with chronic disease in tobacco producing and non-producing districts. Participants were recruited via simple random sampling from local primary care patient registries, with a fixed allocation technique. Data were collected by face-to-face interviews with a verified structured questionnaire. Ordinal logistic regression was used to identify predictors of smoking status. The overall smoking rate was 19.3% (7.5% current and 11.8% former). Males accounted for 98.5% of individuals who had previously smoked. High morbidity was noted, with 28.8% of smokers having COPD and 54.7% requiring at least four hospitalizations per year. Traditional hand-rolled tobacco (Yasen) dominated usage, especially in production regions (88.1% vs. 75.0%). Multivariable analysis found that females had 99.6% lower odds of smoking (AOR: 0.004, 95% CI: 0.002–0.009). Patients from non-planting areas (AOR: 0.47, 95% CI: 0.28–0.77) and low-income households (AOR: 0.50, 95% CI: 0.26–0.95) had considerably decreased smoking odds. While current prevalence is lower than national averages. Geographic closeness to production and higher household income both contribute to smoking in this group. Therefore, Interventions have to address physical and economic accessibility to traditional tobacco in agricultural communities. Introduction Tobacco is a substance made from the leaves of nicotiana plants that contains highly addictive stimulant nicotine. Its use involves smoking or smokeless products, which primarily cause addiction owing to nicotine’s release of dopamine in the brain1,. Globally, over 1.1 billion individuals aged 15 and over smoke, with 942 million males and 175 million women currently smoking2,3. The macroeconomic context and structural environment of this epidemic is very uneven, as over 80% of smokers reside in low- and middle-income nations (LMICs)3. This concentration reflects a huge shift in worldwide agricultural output, with tobacco production and smoking rates shifting from high-income to low- and middle-income countries. Between 1961 and 2023, Asia and Africa’s share of global tobacco leaf output more than doubled, from 44.3% in 1961 to 78.8% in 20234. According to WHO regional reports, Europe has highest prevalence of tobacco smoking5. However, the Asian continent t bears a staggering absolute volume of tobacco consumers. According to a 2021 report, there were roughly 137 million current smokers aged 15 and older in Asia6. South Asia is a global high point for tobacco smoking, accounting for 25% of world’s smokers. The WHO South-East Asia area is home to about 90% of the world’s smokeless tobacco users7,8, Nearly half the world’s smokers (46%) live in China India and Indonesia, where 290 million, 116 million 61 million current smokers respectively9 whereas in Thailand, more than 14.6 million people use tobacco, which causes over 80,000 fatalities per year10,11. Southern Thailand has repeatedly been identified as having the highest smoking rate in the country, with reports indicating rates of 25.6% or higher, with the highest proportion of smokers found in Satun, Pattani, Ranong, Nakhon Si Thammarat, Surat Thani, and Trang12. Tobacco use remains an important behavioral risk factor for noncommunicable diseases, contributing significantly to preventable morbidity in people with chronic conditions. Despite long-standing national tobacco-control efforts, smoking continues to be a major public health risk across vulnerable demographics. A national survey of Thai adults revealed an overall smoking prevalence of 18.2%, with smoking behavior being significantly associated with individual socioeconomic conditions, secondhand smoke exposure, concurrent alcohol consumption, and social marketing factors such as direct exposure to health warnings and anti-smoking information13. This complex behavioral landscape is more prominent in older groups. Previous nationally representative evidence focused especially on Thai older adults found a smoking prevalence of 15.11%14. This data showed that older persons’ smoking behavior was substantially connected with male gender, younger age, working status, physical health status, alcohol use, and community participation14. Taken together, these findings emphasize the importance of investigating smoking behavior in larger social, environmental, and communication contexts, implying that tobacco use in later life is influenced by multiple health determinants and social roles rather than individual preference13,14. Despite a relative fall in the number of smokers, tobacco smoking continues to cause millions of premature deaths each year15. It kills up to half of its users, which equals to 8 million deaths per year globally and is currently the world’s single largest cause of unnecessary mortality16. It accounted for 10·8% of all-cause mortality in the Asian region6. It has a catastrophic influence on health by generating chronic diseases, weakening the immune system, and harming practically every organ in the body17–19. It puts smokers at risk of respiratory tract infections17,20. Smoking causes eight out of ten COPD-related deaths. It causes chronic bronchitis and emphysema, which erode lung tissue and airways, resulting in “smoker’s cough” and diminished ability to breathe21–23. In 2021, tobacco was responsible for 64% of trachea, bronchus, and lung cancer deaths, 16% of ischemic heart disease deaths, and 38% of COPD deaths20,23. It plays an important role in atherosclerosis, which causes cardiovascular disease (CVD), chronic obstructive pulmonary disease (COPD), type 2 diabetes, and numerous malignancies18,19. Despite its involvement in exacerbating diseases such as cancer, diabetes, and heart disease, tobacco use remains prevalent among chronic disease patients, with smoking rates reaching 51.9% in some high-risk populations24. Patients with chronic diseases such as hypertension, diabetes, and COPD have the highest risk for tobacco-related problems25,26. In addition to individual considerations, environmental and spatial elements have a major influence on smoking behavior. Spatial evidence from Thailand suggests that the number of smokers varies regionally and may be connected with area-specific vulnerability to respiratory health hazards, as well as larger health vulnerabilities at the provincial level27,28. This emphasizes the crucial need to investigate smoking behavior not only at the individual level, but also in specific community situations27,29. This structural relationship is especially important in committed tobacco-producing areas. To reduce the health impacts of tobacco use, governments and public health organizations rely on the WHO Framework Convention on Tobacco Control (WHO FCTC) and MPOWER strategies30,31. Thailand has robust national tobacco control policies (taxation, plain packaging)32. However, these rules are largely aimed with commercially made cigarettes. In Southern Thailand, where tobacco is grown locally, “Yasen” (hand-rolled tobacco) is widely available33. These local production facilities often provide citizens with easier access to tobacco at lower-than-market prices and create a strong environment for the accessibility, affordability, and social normalization of tobacco use27,33. At the same time, Thailand’s quick shift to a fully aged society has put significant strain on the public health system, particularly because older persons frequently require ongoing, sophisticated care for chronic noncommunicable diseases34,35. Evidence from provincial-level evaluations under the Universal Coverage Scheme (UCS) shows that health service capacity, service congestion, and broader socioeconomic factors have a significant impact on older individuals’ experiences with the state health-care system34,36. As a result, continuous smoking among patients with chronic conditions may increase healthcare usage and exacerbate the overall stress on public infrastructure34. Examining smoking behavior and related health outcomes in chronic disease patients from tobacco-producing and non-producing locations is therefore critical for determining locally suitable tobacco-control measures and robust health-care system planning. This study looks into whether living in these production sites is significantly connected with increased smoking behavior and adverse health outcomes compared to living in non-producing areas. Methods and materials Study area and period The study was done in two separate districts of Thailand’s Nakhon Si Thammarat Province: Tha Sala District (a designated tobacco-producing area) and Bang Khan District (not a tobacco-producing area). Four sub-districts were chosen from these districts to represent diverse environmental and intervention contexts: Tobacco-Producing Sub-districts: Moklan (smoke-free campaign) and Klai (no campaign). Non-Producing Areas: Wang Hin Sub-district (smoke-free campaign) and Ban Nikhom Subdistrict (no campaign). The data was collected March to May 2019. Study design Analytical cross-sectional study design was employed to collect simultaneous assessment of tobacco-smoking behaviors, health consequences, and quitting methods across the four unique study sites at a single point in time. Population, inclusion, and exclusion criteria The study population is patients who has been diagnosed with at least one of the following non-communicable diseases (NCDs): Chronic Obstructive Pulmonary Disease (COPD), Diabetes Mellitus, or Hypertension for at least 12 months, and resided in the four target subdistricts for at least 12 months. Patients with mental impairment or dementia or in a state of medical emergency were excluded from the study. Sample size determination and sampling procedure The formula for comparing two independent proportions (P1 vs. P2) was used to calculate the sample size. It was estimated based on the assumption that 40.5% of NCD patients smoked (P1)37. According to the study’s hypothesis, prevalence would be 10% higher (P2 = 50.5%) in tobacco-producing areas38. For this sample size calculation, α = 0.05 and a power of 80% were used. Accordingly, 389 participants per district were required. Nevertheless, considering a 13% non-response rate39, tobacco producing area 440 and tobacco non-producing area (439) with a total of 879 of sample size. To ensure balanced groups for the comparative analysis, study participants were chosen using a fixed allocation technique stratified by geographic region (tobacco-producing vs. non-producing districts). To ensure that every eligible individual had a known, probability of selection, the official primary care patient registers controlled by local sub-district health promotion facilities were used as the definitive, closed sample frame. Each district registry collected a detailed census of eligible chronic disease patients who met the inclusion criteria. A computer-generated basic random sampling lottery was then conducted using these fixed register listings until the desired quota of 440 participants per district was met. The total number of eligible chronic illness patients in the registry database was finite and known at baseline (N), hence the probability (Pr) of selection for each particular patient was uniform and mathematically defined as Pr = 440 / N. Variables Dependent variables Tobacco-smoking behavior (Current vs. Former/Never) and adverse health outcomes. Independent v ariables: Primary: Living area (Producing vs. Non-producing), Socio-demographic: Age, gender, marital status, education level, personal and household income, employment status. Behavioral: Age of smoking initiation, quit methods. Operational definitions Current smoker: someone who has smoked within the last 28 days and has smoked at least 100 cigarettes in their lifetime. Former smoker: someone who hasn’t smoked in the last 28 days but has smoked at least 100 cigarettes throughout their lifetime. Never smoked: A person who has never smoked or who has smoked fewer than 100 cigarettes in their lifetime is considered a never smoker40,41:. Adverse health outcomes are operationally defined and measured as the presence of specific self-reported or clinically confirmed respiratory conditions and symptoms like chronic cough, dyspnea, COPD and others. Tool development, data collection method and data quality control A systematic, closed-ended questionnaire developed following a thorough analysis of health economic effect studies and the Global Adult Tobacco Survey (GATS). The tool comprised sociodemographic variables, smoking history and practices, quitting techniques, and smoke-related complications. To ensure content validity, a panel of three independent clinical medicine and public health experts reviewed the tool’s first draft. The experts used a 4-point ordinal relevance scale to score each questionnaire item. An item-level Content Validity Index (I-CVI) was calculated as the percentage of experts who gave a rating of 3 or 4. those with an I-CVI of 1.00 were kept unchanged, whereas those with low scores were progressively altered or removed to improve cultural relevance and clinical precision. A pilot field test was undertaken with thirty people recruited from a demographically similar sub-district in Nakhon Si Thammarat Province, however they were entirely excluded from the final analytical sample. This pilot test evaluated questionnaire flow, language clarity, cross-cultural understanding, and interview duration. Minor language changes were made based on participant comments to avoid cognitive strain during face-to-face administration. Cronbach’s alpha was used to statistically analyze construct internal consistency and reliability based on data from the pilot test. The instrument showed great dependability, with alpha values of 0.8. Trained data collectors conducted face-to-face interviews in order to collect data. For one day, collectors received training on handling data, rules of ethics, and specific objectives of the study. To protect patients’ privacy and comfort, interviews were held at community centers or at their homes. Every day, supervisors verified the uniformity and thoroughness of the surveys. Data collectors were told to explain questions to participants without influencing their responses. To reduce errors, a double-entry data approach was employed at the digital input stage. Ethical considerations This study followed the Declaration of Helsinki and gained formal ethical approval from Walailak University, protocol ID WUEC-18-094-01. After being made aware of an objective of the study and their right to withdraw, participants gave their informed verbal and/or written consent. Unique identifying codes were used to anonymize the identities of the participants. There were no biological samples taken. Data processing, management and analysis The data was exported to SPSS version 26.0 for further analysis, once its completeness and accuracy were assessed. The sociodemographic and smoking profiles were summarized using descriptive techniques such as percentages and frequencies. Tables and text were used to present the results. Using definitions from the U.S. CDC, National Health Interview Survey, and New Zealand Ministry of Health, smoking status was divided into three categories: current smoker, past smoker, and never smoker40,41: Current smoker = Life-time smoked 100 + cigarettes and smoked in the last 28 days.former smoker = Life-time smoked 100 + cigarettes and did not smoke in the last 28 days.never smoker = Life-time smoked < 100 cigarettes. Current smoker = Life-time smoked 100 + cigarettes and smoked in the last 28 days. Current smoker = Life-time smoked 100 + cigarettes and smoked in the last 28 days. former smoker = Life-time smoked 100 + cigarettes and did not smoke in the last 28 days. former smoker = Life-time smoked 100 + cigarettes and did not smoke in the last 28 days. never smoker = Life-time smoked < 100 cigarettes. never smoker = Life-time smoked < 100 cigarettes. Ordinal logistic regression was performed to identify determinants of smoking status among study participants. A purposeful variable selection method was used to build the multivariable model. Candidates for multivariate ordinal logistic regression analysis were initially screened and found in bivariate analysis using cutoff criterion of p-values < 0.25. To control for confounding, the selected candidates were entered into the multivariable regression model at the same time. Multivariate analysis identified variables with p-values < 0.05 as significant predictors of tobacco smoking behavior. The strength of associations was declared using adjusted odds ratios (AOR) and a 95% confidence interval (95%CI). Results Socio-demographic and economic characteristics A total of 879 chronic disease patients completed the study with a 100% response rate; employing a fixed allocation technique, this included 440 participants (50.1%) from the tobacco-producing area and 439 participants (49.9%) from the non-producing area. Only 25.75% of the participants in this study were men, compared to roughly 74.3% who were women. The participants were 64.9 years old on average. Just 1.4% of the participants have a bachelor’s degree, compared to 88.6% who have only completed grade 6 or less. Approximately 30.6% of the participants live in low-income households whereas only 17.6% of the participants are in high-income category. Regarding monthly personal income, about 38.5% of the study participants fell into the lower income quartile while 18% are at the upper quartile. Furthermore, 41.9% of them work in agriculture or as daily laborers, but more than half (48.6%) are retired, unemployed, or stay at home. 95.2% of the members rely solely on the state-funded Universal Health Coverage (“Gold Card”) program (Table 1). Table 1Socio-demographic and economic characteristics of study participants.VariablesCategoriesFrequency (N = 879)Percent (%)SexMale22625.7Female65374.3ResidencesTobacco producing area44050.1Tobacco non-producing area43949.4Marital statusMarried66275.3Divorced16619.4Single293.3Widowed182Household income (missing = 4)Low income26830.6Middle income21824.9Upper middle income23526.9High income15417.6Age (years)Mean (SD)64.911.5Age categories in years18–5429133.155–6932737.270+26129.7Education status (missing = 2)Grade 6 or lower77788.6Grade 7–128810.0Bachelor degree121.4Health insurance schemeUniversal Health Coverage83795.2Social Security50.6Civil Servants242.7Others50.6Do not know80.9Personal monthly income in Thai baht (missing = 1)Lower quartile33838.5Second quartile22125.2Third quartile16118.3Upper quartile15818Personal income/month (Thai baht) (missing = 1)Median (Q1, Q3)2500 (700, 5000)Occupation (missing = 4)Agriculture/daily labor36741.9Civil Servants/State enterprise20.2Private sector (monthly salary)20.2Owning Business799.0Retired/unemployed/house making42548.6Self-rated health status (missing = 4)Very good414.7Good25228.8Moderate45251.7Bad12714.5Very bad30.3 Table 1Socio-demographic and economic characteristics of study participants.VariablesCategoriesFrequency (N = 879)Percent (%)SexMale22625.7Female65374.3ResidencesTobacco producing area44050.1Tobacco non-producing area43949.4Marital statusMarried66275.3Divorced16619.4Single293.3Widowed182Household income (missing = 4)Low income26830.6Middle income21824.9Upper middle income23526.9High income15417.6Age (years)Mean (SD)64.911.5Age categories in years18–5429133.155–6932737.270+26129.7Education status (missing = 2)Grade 6 or lower77788.6Grade 7–128810.0Bachelor degree121.4Health insurance schemeUniversal Health Coverage83795.2Social Security50.6Civil Servants242.7Others50.6Do not know80.9Personal monthly income in Thai baht (missing = 1)Lower quartile33838.5Second quartile22125.2Third quartile16118.3Upper quartile15818Personal income/month (Thai baht) (missing = 1)Median (Q1, Q3)2500 (700, 5000)Occupation (missing = 4)Agriculture/daily labor36741.9Civil Servants/State enterprise20.2Private sector (monthly salary)20.2Owning Business799.0Retired/unemployed/house making42548.6Self-rated health status (missing = 4)Very good414.7Good25228.8Moderate45251.7Bad12714.5Very bad30.3 Socio-demographic and economic characteristics of study participants. Smoking profile, health outcomes and health care utilization From the study population with non-communicable chronic diseases about 80.7% had no history of smoking. However, the 19.3% of the participants were with the history of tobacco use containing both current and former smokers. The median age for smoking initiation was 18 years while the mean age this study subjects is approximately 65 years. While 24.8% of the overall study population has dyspnea, this increases dramatically to 37.6% among individuals with a smoking history. Similarly, the smoking group has a higher rate of chronic cough (18.2%) than the general study group (13.7%). Smokers have roughly equal rates of productive cough (36.5%) and nonproductive cough (34.7%). While COPD prevalence is much greater in the smoking history group (28.8%) than in the general population (7.3%). Moreover, study subjects reported (35.9%) that pneumonia or bronchitis as acute secondary infections. More than half (54.7%) of the COPD patients required four or more hospital admission within the last 12 months. In addition, emphysema is the most common smoking-related consequence (37.6%), followed by acute secondary infections such as pneumonia and bronchitis (35.9%). Regarding comorbidity profile of individuals with history of tobacco use, about 71.8%, 35.9% and 25.9% suffer from hypertension, dyslipidemia and diabetes mellitus respectively (Table 2). Table 2Tobacco smoking behavioral and health outcome among chronic disease patients.VariablesCategories(N = 879)%Smoking statusNever smoker70980.7Former smoker10411.8Current smoker667.5Age at first smoking (years)Median (Q1, Q3)18 (15, 20)Dyspnea symptoms (n = 879)Yes21824.8No66175.2Dyspnea symptoms (n = 170)Yes6437.6No10662.4Chronic cough (n = 879)Yes12013.7No75986.3Chronic cough (former smoker vs. current smoker) (n = 170)Yes3118.2No13981.8Type of cough (n = 170)Productive6236.5Non-productive5934.7Adverse health outcomes /Smoke related complication (n = 170)Emphysema6437.6Secondary infection (pneumonia and bronchitis)6135.9Atherosclerosis & Arterial Stiffness158.8COPD (2 missing)Yes647.3No81392.7COPD (n = 170)Yes4928.8No12171.2Frequency of admission from COPD in past 12 months (n = 64)1–3 times2742.2≥ 4 times3554.7Diabetes Mellitus (n = 879)Yes35740.6No52259.4DM (n = 170)Yes4425.9No12674.1Hypertension (n = 879)Yes73884No14116Hypertension (n = 170)Yes12271.8No4828.2Dyslipidemia (n = 879)Yes54361.8No33638.2Dyslipidemia (n = 170)Yes6135.9No10964.1Ischemic Heart diseases (n = 879)Yes232.6No85697.4Ischemic Heart diseases (n = 170)Yes63.5No16496.5Others complications (health failure, cancer) (n = 879)Yes252.8No85497.2Others complications (health failure, cancer) (n = 170))Yes52.9No16597.1 Table 2Tobacco smoking behavioral and health outcome among chronic disease patients.VariablesCategories(N = 879)%Smoking statusNever smoker70980.7Former smoker10411.8Current smoker667.5Age at first smoking (years)Median (Q1, Q3)18 (15, 20)Dyspnea symptoms (n = 879)Yes21824.8No66175.2Dyspnea symptoms (n = 170)Yes6437.6No10662.4Chronic cough (n = 879)Yes12013.7No75986.3Chronic cough (former smoker vs. current smoker) (n = 170)Yes3118.2No13981.8Type of cough (n = 170)Productive6236.5Non-productive5934.7Adverse health outcomes /Smoke related complication (n = 170)Emphysema6437.6Secondary infection (pneumonia and bronchitis)6135.9Atherosclerosis & Arterial Stiffness158.8COPD (2 missing)Yes647.3No81392.7COPD (n = 170)Yes4928.8No12171.2Frequency of admission from COPD in past 12 months (n = 64)1–3 times2742.2≥ 4 times3554.7Diabetes Mellitus (n = 879)Yes35740.6No52259.4DM (n = 170)Yes4425.9No12674.1Hypertension (n = 879)Yes73884No14116Hypertension (n = 170)Yes12271.8No4828.2Dyslipidemia (n = 879)Yes54361.8No33638.2Dyslipidemia (n = 170)Yes6135.9No10964.1Ischemic Heart diseases (n = 879)Yes232.6No85697.4Ischemic Heart diseases (n = 170)Yes63.5No16496.5Others complications (health failure, cancer) (n = 879)Yes252.8No85497.2Others complications (health failure, cancer) (n = 170))Yes52.9No16597.1 Tobacco smoking behavioral and health outcome among chronic disease patients. Tobacco smoking status, cessation behavior and respiratory health outcomes in tobacco-producing and non-producing areas While the frequency of former smokers ranges from 9.7% to 14.2%, the prevalence of current smokers ranges from 4.1% to 9.7%, with the highest incidence observed in a tobacco-producing area, as demonstrated in the Klai sub-district. This indicates that a higher proportion of people are ex-smokers. The average age at which people start smoking is 16–18 years old, and this is true for all four subdistricts, including those that do not produce tobacco. In all four sub-districts, the majority of participants (92.7% to 98.2%) said they would rather quit by themselves. Only the use of nicotine replacement therapy or smoke cessation clinics is modest, though. Between 75% and 88.1% of tobacco consumption is either Yasen (hand-rolled tobacco) or Bai-Jak (leaf-wrapped). The percentage of hand-rolled tobacco is higher in tobacco-producing regions (87.5%–88.1%) than in non-producing regions (75%–76.1%). Shortness of breath is the most prevalent respiratory symptom, affecting between 21% and 29.9% of study subjects. Tobacco-producing areas have the greatest rates of shortness of breath (27.2% in Klai and 29.9% in Moklan). The percentage of participants with smoking-related problems ranged from 9.6% to 14.9% (Table 3). Table 3Smoking Status, Cessation Behaviors, and Respiratory Health Outcomes between Tobacco-Producing and Non-Producing Areas (n = 170).variablesTobacco-producing areasNon-producing areasKlaiMoklanBan NikhomWang HinSmoking status (%)Former smoker9.711.714.211.8Current smoker9.77.94.18.3Age in years (median) at first smoke18161818Quite method (%)Smoking cessation clinics in healthcare facilities10.82.70.07.3Purchased Nicotine Replacement Therapy (NRT)0.00.02.610.0Used alternative medicine0.00.02.60.0Quit on my own/self-quit98.297.394.992.7Type of most recent cigarette use (%)High-price Domestic Cigarettes10.011.920.017.4Budget Domestic Cigarettes/lower-tier Thai Tobacco Board brands2.50.05.04.3Budget Imported Cigarettes/low-cost0.00.00.02.2Hand-rolled Tobacco / Leaf-wrapped Tobacco87.588.175.076.1Dyspnea (%)27.229.921.521.0Chronic cough (%)14.312.115.113.1Smoking related complications (%)14.912.69.69.6 Table 3Smoking Status, Cessation Behaviors, and Respiratory Health Outcomes between Tobacco-Producing and Non-Producing Areas (n = 170).variablesTobacco-producing areasNon-producing areasKlaiMoklanBan NikhomWang HinSmoking status (%)Former smoker9.711.714.211.8Current smoker9.77.94.18.3Age in years (median) at first smoke18161818Quite method (%)Smoking cessation clinics in healthcare facilities10.82.70.07.3Purchased Nicotine Replacement Therapy (NRT)0.00.02.610.0Used alternative medicine0.00.02.60.0Quit on my own/self-quit98.297.394.992.7Type of most recent cigarette use (%)High-price Domestic Cigarettes10.011.920.017.4Budget Domestic Cigarettes/lower-tier Thai Tobacco Board brands2.50.05.04.3Budget Imported Cigarettes/low-cost0.00.00.02.2Hand-rolled Tobacco / Leaf-wrapped Tobacco87.588.175.076.1Dyspnea (%)27.229.921.521.0Chronic cough (%)14.312.115.113.1Smoking related complications (%)14.912.69.69.6 Smoking Status, Cessation Behaviors, and Respiratory Health Outcomes between Tobacco-Producing and Non-Producing Areas (n = 170). Health outcomes by smoking status and geographic area Pearson’s Chi-square test was used across the study population (N = 879) to investigate how behavioral smoking status and geographic location influence health outcomes. The study found a statistically significant association between smoking behavior and COPD diagnosis (x2 = 165.744, p < 0.001), as well as a significant relationship with active dyspnea (x2 = 18.093, p < 0.001). Notably, a clear sick quitter’ impact was identified across all measures; ex-smokers consistently had the highest clinical and symptomatic burden, with a 33.9% COPD prevalence and a 40.0% dyspnea rate. Chronic cough, with the highest prevalence among ex-smokers (20.9%) was identified. Residing in a tobacco-producing area was substantially related with greater rates of dyspnea than non-producing areas (28.7% vs. 21.0%; x2 = 6.938, p = 0.008) (Table 4). Table 4Health Outcomes by Smoking Status and Geographic Area (N = 879).Health OutcomesSmoking StatusGeographic AreaNever Smoker(n = 695) (%)Ex-Smoker (n = 115) (%)Current Smoker(n = 69) (%)Pearson x2p-valueNon-Producing Area (n = 439) (%)Tobacco-Producing Area (= 440) (%)x2, p-valueDyspnea18.093< 0.0016.938, 0.008Yes152 (21.9%)46 (40.0%)20 (29.0%)92 (21.0%)126 (28.7%)No543 (78.1%)69 (60.0%)49 (71.0%)347 (79.0%)314 (71.3%)Chronic Cough5.8620.0530.037, 0.847Yes87 (12.5%)24 (20.9%)9 (13.0%)61 (13.9%)59 (13.4%)No608 (87.5%)91 (79.1%)60 (87.0%)378 (86.1%)381 (86.6%)COPD Clinical Diagnosis165.744< 0.0010.884, 0.347Yes12 (1.7%)39 (33.9%)13 (18.8%)29 (6.5%)36 (8.1%)No681 (98.3%)76 (66.1%)56 (81.2%) Table 4Health Outcomes by Smoking Status and Geographic Area (N = 879).Health OutcomesSmoking StatusGeographic AreaNever Smoker(n = 695) (%)Ex-Smoker (n = 115) (%)Current Smoker(n = 69) (%)Pearson x2p-valueNon-Producing Area (n = 439) (%)Tobacco-Producing Area (= 440) (%)x2, p-valueDyspnea18.093< 0.0016.938, 0.008Yes152 (21.9%)46 (40.0%)20 (29.0%)92 (21.0%)126 (28.7%)No543 (78.1%)69 (60.0%)49 (71.0%)347 (79.0%)314 (71.3%)Chronic Cough5.8620.0530.037, 0.847Yes87 (12.5%)24 (20.9%)9 (13.0%)61 (13.9%)59 (13.4%)No608 (87.5%)91 (79.1%)60 (87.0%)378 (86.1%)381 (86.6%)COPD Clinical Diagnosis165.744< 0.0010.884, 0.347Yes12 (1.7%)39 (33.9%)13 (18.8%)29 (6.5%)36 (8.1%)No681 (98.3%)76 (66.1%)56 (81.2%) Health Outcomes by Smoking Status and Geographic Area (N = 879). Factors associated with smoking status among chronic patients In bivariate analysis, about seven candidate variables were identified for the multivariate ordinal logistic regression analysis based on the p-values of < 0.25 as a cutoff point. Candidates were gender, household income, marital status, employment, age, personal income and tobacco planting area. In multivariate ordinal logistic regression analysis, three independent variables; gender, tobacco planting area and household income were identified based on a p-value < 0.05 cut-off points. The multivariable ordinal logistic regression model constructed through conscious variable selection revealed a strong overall statistical fit. The final model with joint predictors beats the baseline intercept-only model (X2 = 535.44, df = 18, p < 0.001), indicating good explanatory capability based on socio-demographic and geographic data. The entire model fit was further reinforced by a highly satisfactory, non-significant Deviance goodness-of-fit estimate (X2 = 413.62, df = 940, p = 1000). The Pearson goodness-of-fit test was significant (X2=1231.12, df = 940, p < 0.001), but this is a known mathematical artifact in large datasets with multi-categorical predictors that generate sparse contingency matrix cell counts. Therefore, the Deviance measure is a more stable indicator of model calibration here. Additionally, the model showed strong explanatory value with a Nagelkerke pseudo-R2 of 0.649. The fundamental proportional odds assumption was fully met, as the test of parallel lines produced a highly non-significant result (X2 = 8.78, df = 18, p = 0.965), confirming that the location parameters and slope coefficients remain uniform across the cumulative logit response boundaries. Importantly, a rigorous investigation of potential two-way interaction terms revealed no statistically significant interaction effects, particularly between Geographic Area and Household Income (Wald x2 = 1.37, df = 2, p = 0.504). This suggests that the effect of living in a tobacco-producing area on smoking status is consistent and irrespective of a patient’s household income level. According to the study’s findings, females are 99.6% less likely than men to be in a higher risk category (ex-smoker or current smoker) (AOR: 0.004, CI: 0.002–0.009). Compared to those who do not live in the tobacco planting area, patients who do so had 53% lower odds of smoking (AOR: 0.47, CI: 0.28–0.77). Furthermore, compared to those from high-income families, patents from low-income households have 50% reduced risks of smoking (AOR: 0.50, 95% CI: 0.26–0.95) (Table 5). Table 5Factors associated with smoking status among chronic patientsVariablesCategoriesNever Smoker n (%)Ex-Smoker n (%)Current Smoker n (%)COR (95% CI)p-valueAOR (95% CI)p-valueGenderMale63(27.9%)98(43.4%)65(28.8%)1.00 (Ref)--1.00 (Ref)--Female646(98.9%)6(0.9%)1 (0.2%)0.004 (0.001–0.02)< 0.0010.004 (0.002–0.009)< 0.001*Tobacco areaPlanting347(7%)64(14.6%)28(6.4%)1.00 (Ref)--1.00 (Ref)--Non-Planting362(82.3%)40(9.1%)38(8.6%)0.93(0.67–1.30)0.680.47 (0.287–0.769)0.003*Household incomeHigh118(70.2%)26(15.5%)24(14.3%)1.00 (Ref)--1.00 (Ref)--Middle295(83.6%)31(8.8%)27(7.6%)0.46 (0.30–0.70)< 0.0010.67 (0.37–1.22)0.207Low293(82.8%)47(13.3%)14(4.0%)0.46 (0.30–0.71)< 0.0010.50 (0.259–0.951) 0.035 Age groups in years70+263(79.7%)47(14.2%)20(6.1%)1.00 (Ref)--1.00 (Ref)--55–69296(78.1%)49(12.9%)34(9.0%)1.13 (0.79–1.62)0.5050.87 (0.55–1.38)0.61318–54150(88.2%)8(4.7%)12(7.1%)0.55 (0.32–0.94)0.0290.77 (0.33–1.79)0.397EducationBachelor’s Degree9(75.0%)1(8.3%)2(16.7%)1.00 (Ref)--1.00 (Ref)--(Grades 7–12)64(72.7%)13(14.8%)11(12.5%)1.05(0.27–4.02)0.95(Grade 6 or Lower)634(81.6%)90(11.6%)53(6.8%)0.62 (0.17–2.21)0.458EmploymentAgriculture and Daily Labor275(74.9%)50(13.6%)42(11.4%)1.86(1.30– 2.62)0.001*1.32(0.67– 2.62)0.418Business Owners68(86.1%)7(8.9%)4 (5.1%)0.88(0.44– 1.74)0.7050.86(0.29– 2.53)0.778Non-Active (retired, unemployed)358(84.2%)47(11.1%)20(4.7%)1.00 (Ref)--1:00 (Ref)---CampaignNo354(81.4%)51(11.7%)30(6.9%)1.00 (Ref)------Yes355 (80.0%)53(11.9%)36(8.1%)1.12 (0.76–1.63)0.563----Personal incomeHigh404(79.2%)61(12.0%)45(8.8%)1.00 (Ref)--1:00 (Ref)--Low304(82.6%)43(11.7%)21(5.7%)0.78 (0.55–1.10)0.1701.6(0.79–3.24)0.189Marital statusMarried513(77.5%)93(14.0%)56(8.5%)3.16(1.77–5.63)< 0.001*10.4(0.42– 2.54)0.933Divorced152(91.6%)10(6.0%)4 (2.4%)1:00 (Ref)--1:00 (Ref)--Single25(86.2%)1(3.4%)3(10.3%)1.86(0.58– 5.97)0.2961.96(0.37 − 10.41)0.429Widowed16(88.9%)0(0.0%)2(11.1%)1.48(0.32–6.81)0.6110.75(0.11–5.24)0.776Monthly personal income high (median > 2500 baht) & low (median < 2500 baht), household income. Table 5Factors associated with smoking status among chronic patientsVariablesCategoriesNever Smoker n (%)Ex-Smoker n (%)Current Smoker n (%)COR (95% CI)p-valueAOR (95% CI)p-valueGenderMale63(27.9%)98(43.4%)65(28.8%)1.00 (Ref)--1.00 (Ref)--Female646(98.9%)6(0.9%)1 (0.2%)0.004 (0.001–0.02)< 0.0010.004 (0.002–0.009)< 0.001*Tobacco areaPlanting347(7%)64(14.6%)28(6.4%)1.00 (Ref)--1.00 (Ref)--Non-Planting362(82.3%)40(9.1%)38(8.6%)0.93(0.67–1.30)0.680.47 (0.287–0.769)0.003*Household incomeHigh118(70.2%)26(15.5%)24(14.3%)1.00 (Ref)--1.00 (Ref)--Middle295(83.6%)31(8.8%)27(7.6%)0.46 (0.30–0.70)< 0.0010.67 (0.37–1.22)0.207Low293(82.8%)47(13.3%)14(4.0%)0.46 (0.30–0.71)< 0.0010.50 (0.259–0.951) 0.035 Age groups in years70+263(79.7%)47(14.2%)20(6.1%)1.00 (Ref)--1.00 (Ref)--55–69296(78.1%)49(12.9%)34(9.0%)1.13 (0.79–1.62)0.5050.87 (0.55–1.38)0.61318–54150(88.2%)8(4.7%)12(7.1%)0.55 (0.32–0.94)0.0290.77 (0.33–1.79)0.397EducationBachelor’s Degree9(75.0%)1(8.3%)2(16.7%)1.00 (Ref)--1.00 (Ref)--(Grades 7–12)64(72.7%)13(14.8%)11(12.5%)1.05(0.27–4.02)0.95(Grade 6 or Lower)634(81.6%)90(11.6%)53(6.8%)0.62 (0.17–2.21)0.458EmploymentAgriculture and Daily Labor275(74.9%)50(13.6%)42(11.4%)1.86(1.30– 2.62)0.001*1.32(0.67– 2.62)0.418Business Owners68(86.1%)7(8.9%)4 (5.1%)0.88(0.44– 1.74)0.7050.86(0.29– 2.53)0.778Non-Active (retired, unemployed)358(84.2%)47(11.1%)20(4.7%)1.00 (Ref)--1:00 (Ref)---CampaignNo354(81.4%)51(11.7%)30(6.9%)1.00 (Ref)------Yes355 (80.0%)53(11.9%)36(8.1%)1.12 (0.76–1.63)0.563----Personal incomeHigh404(79.2%)61(12.0%)45(8.8%)1.00 (Ref)--1:00 (Ref)--Low304(82.6%)43(11.7%)21(5.7%)0.78 (0.55–1.10)0.1701.6(0.79–3.24)0.189Marital statusMarried513(77.5%)93(14.0%)56(8.5%)3.16(1.77–5.63)< 0.001*10.4(0.42– 2.54)0.933Divorced152(91.6%)10(6.0%)4 (2.4%)1:00 (Ref)--1:00 (Ref)--Single25(86.2%)1(3.4%)3(10.3%)1.86(0.58– 5.97)0.2961.96(0.37 − 10.41)0.429Widowed16(88.9%)0(0.0%)2(11.1%)1.48(0.32–6.81)0.6110.75(0.11–5.24)0.776Monthly personal income high (median > 2500 baht) & low (median < 2500 baht), household income. Factors associated with smoking status among chronic patients Monthly personal income high (median > 2500 baht) & low (median < 2500 baht), household income. Discussion This study was aimed to determine prevalence of tobacco uses among chronic patients living both in tobacco-producing and non-producing area of the two distinct districts in Nakhon Si Thammarat Province, Thailand. Accordingly, overall prevalence was 19.3% while tobacco smoking prevalence among former and current smokers were11.8% and 7.5% respectively. In this study, the overall tobacco smoking prevalence was 19.3% (7.5% current and 11.8% former) among chronic diseases patients. Current smoking prevalence is 7.5% which is lower than 12% reported in Brazil’s Hiperdia Minas42, 11.8% reported in Iranian study43, 19.78% reported in Shanghai, China44, 21.08% reported in Hebei Province, Central China45, the 26.2% reported among chronically ill patients in rural Western China46, the 40.3% current use in Kathmandu, Nepal47, and the 15.5% reported in the Cambodia, Myanmar, and Vietnam study24. The current smoking prevalence in the present study was lower, possibly because the participants were patients with established chronic diseases and may have reduced or stopped smoking after diagnosis. While overall smoking prevalence (19.3%) is higher than a study conducted chronic diseases patients in Thailand, 12%48, which is comparable to the national smoking prevalence of 18.2% reported in a large Thai survey13. However, this prevalence is lower than global monitoring target (29.2%)49, 33.7% projected for the Thai population under a ‘Combined Policies’ scenario for 202550. The disparities in smoking prevalence among studies could be attributed to population characteristics, diseases status, and local tobacco availability may partly explain the observed variation between the present finding and national-level estimate towards smoking behavior13. These findings are also influenced by the study subjects’ socio-demographic makeup, specifically the high proportion of female participants (74.3%) in a cultural environment where female smoking is uncommon, as well as the older age groups. The environmental availability of traditional hand-rolled tobacco (Yasen) in producing regions explains why tobacco usage has remained greater than in non-agricultural provinces. The differences might be also due to the variations in the study period, data collection methods and sample sizes. Regarding respiratory distress, in this study about 24.8% of dyspnea among total chronic diseases patients and 37.6% of dyspnea among smokers were identifies. This finding is significantly lower than the 65.2% in s study conducted in Jordan51 and the 70.6%-72.5% reported in the study in Spain52. While regarding chronic cough, 13.7% among total chronic disease patients and 18.2% among smokers reported. This finding is lower than the 15% chronic bronchitis rate in the Jordan51 and 61.5% cough prevalence in the study in Spain52. The lower prevalence in this study could be attributed to the high proportion of never-smokers (80.7%), as well as the fact that many smokers used traditional hand-rolled tobacco (Yasen), which may have different smoking patterns or symptom reporting thresholds than commercially available cigarette users in Jordan or Spain. This study found that chronic care patients residing in tobacco-producing areas had a considerably higher burden of health outcomes such as dyspnea than those in non-producing areas (28.7% vs. 21.0%; X2 = 6.938, p = 0.008). while acute functional respiratory distress (dyspnea) was highly sensitive to geographic location, long-term structural pathologies such as chronic cough (13.4% vs. 13.9%, p = 0.847) and formal clinical diagnoses of COPD (8.1% vs. 6.5%, p = 0.347) were statistically homogeneous across both regional cohorts. This shows that the initial physiological toll of living in a tobacco-cultivating ecosystem is predominantly manifested as increased acute symptomatic distress rather than immediate changes in persistent, structurally diagnostic clinical outcomes53–55. In addition, about 7.3% of the total chronic disease patients and 28.8% smokers reported COPD. This overall COPD prevalence (7.3%) remarkable consistent with the 8.2% found in Jordan56, and falls between the 4.7% in the Rotterdam Study57, and the 10.1%–13.3% reported in study52. While prevalence of COPD among smokers was 28.8%, which is higher than 24.3% of COPD developed in a continuous smoker in Copenhagen, Denmark58. But lower than 50% probability of developing COPD in lifelong smokers in developed countries59. Moreover, in this study more than 50% of the COPD patients required four or more hospital admission within the last 12 months of prior data collection period of this study. This is consistent with the World Health Organization’s60, suggestions that tobacco is responsible for 30–40% of cases in low- and middle-income countries, although household air pollution is a substantial risk factor. This high proportion of COPD and repeated hospitalization among smokers in the present study should be taken not only as an individual clinical consequence but also as a health system concern. Prior evidence from Thailand has shown that service congestion, bed occupancy, physician availability, and socioeconomic conditions shape older adults’ satisfaction with public health services. Thus, reducing tobacco use among chronic disease patients may contribute to lowering preventable service utilization and improving the responsiveness of the health system34. Regarding comorbidity profile of individuals with history of tobacco use, about 71.8%, 35.9% and 25.9% suffer from hypertension, dyslipidemia and diabetes mellitus respectively. In this study, about 71.8% of the individuals with history of tobacco use had comorbidity profile of hypertension which is lower than the 42.63% reported in General population61. This shows chronic disease and tobacco use create higher risk of increased hypertension than in general population. While about 35.9% of dyslipidemia was identified among individuals with the history of tobacco use, which is slightly higher than 32.8% identified in general population61 and 38% in Iran among non-smokers43. However, dyslipidemia was lower than the 54.9% reported for current smokers in the Iranian study43. Furthermore, 25.9% of diabetes mellitus was identified in this study, which is nearly consistent with the 21.37% reported in the general population61. Between 75% and 88.1% of tobacco consumption is either Yasen (hand-rolled tobacco) or Bai-Jak (leaf-wrapped). The percentage of hand-rolled tobacco is higher in tobacco-producing regions (87.5%–88.1%) than in non-producing regions (75%–76.1%). Which is higher as compared to the 28.4% reported in UK and 6.7% reported in USA62, 24.2% current RYO in Canada63 and 20.1% in six communities of Dominican Republic64. However, 10% to 11.9% use commercial cigarettes in the study area, which is much lower than 58% reported in six economically disadvantaged communities in the Dominican Republic64. Hand-rolled tobacco’s exceptional domination in Nakhon Si Thammarat is primarily due to its economic accessibility, as locally untaxed crops provide a nearly free product for rural communities. This inclination is strongly fostered by cultural normalization, in which traditional types such as Yasen and Bai-Jak are regarded as a typical social norm rather than a niche economic choice. Furthermore, a major regulatory gap exists since these traditional products frequently avoid national “Combined Policies,” such as high taxing and simple packaging, which effectively limit commercial brands. The present study found that smoking history remained common among chronic disease patients, with a particularly strong gender difference. Female patients had substantially lower odds of being in a higher smoking-status category than male patients. Females had 99.6% lower probability of being in a higher risk group than males (AOR:0.004). Furthermore, 28.8% of males are current smokers, while smoking among females is nearly non-existent (0.2%). This finding is consistent with national evidence among Thai older adults, which identified male sex as one of the strongest determinants of smoking behavior. The strong gender gradient observed in both studies may reflect the sociocultural acceptability of smoking among men, gendered patterns of tobacco initiation, and possible underreporting of smoking among women in Thai communities14. Also this finding shows mirror images of WHO report65, that indicates a 36.7% prevalence for men versus 7.8% for women globally. Whereas in Western pacific15, men have greater prevalence (43.3%) but women remain low (2.5%) which is closest to regional parallel to this study. However, prevalence of smoking among females is higher than studies conducted in Canada66 and Europe15, where smoking rates are 26.4% and 17.4% respectively. According to certain comparative analysis studies, women’s tobacco use takes many distinct forms. Men smoke cigarettes while women often consume smokeless tobacco (10.1% prevalence)67. Despite their lower prevalence, females are more vulnerable to the effects of tobacco use, with a 3.3 times higher risk of respiratory problems when they smoke compared to a 1.6 risk for men68. These variations might be due to the variations in socio-cultural differences and reporting bias, where living in a strict cultural norm may affect females’ reporting as compared to other study setting like Canada or Europe. Chronic care patients in tobacco non-producing areas had 53% lower odds of being in a higher smoking risk category (AOR: 0.47, p = 0.003) compared to those in tobacco producing areas. And also this finding aligns with a generalized relationship where tobacco farming leads to a higher risk of use in69–71 and extreme smoking prevalence in China’s top tobacco-producing provinces (68.0% in Yunnan)72. The much greater smoking risk in tobacco-producing areas is due to a unique combination of economic, cultural, and regulatory variables that produce a risky environment. This finding reinforces the importance of area-level context in shaping smoking behavior. Previous spatial research in Thailand demonstrated that smoking prevalence had a geographic pattern and was associated with broader health vulnerability at the provincial level. Therefore, tobacco-producing communities may require spatially targeted tobacco-control strategies rather than uniform national interventions alone27,28. Patients from low-income households had 50% lower odds of smoking (AOR: 0.50, p = 0.035) compared to high-income households. This finding is more consistent with the European adolescent study73, which discovered that personal wealth is positively related to smoking. The majority of studies identified smoking as a disease of poverty, but the results of this study suggest the reverse. This study’s findings directly contradict the study74, which suggests a dose-response relationship in which lower income is related with increased smoking rates. While the study75–80, indicated that 47.9% of the poorest males smoke, compared to only 29.1% of the richest, the findings in this study area show that rich people are twice as likely to be in a higher smoking risk category than the poor. Unlike some other research where smoking is more prevalent among the poor, lower-income patients in this study may face a double burden of NCD treatment expenses and tobacco pricing, and may be forced to quit or remain nonsmokers due to financial restrictions. Conclusion This study found that tobacco use is a persistent problem among chronic disease patients in Nakhon Si Thammarat, Thailand, with an overall prevalence of 19.3%. The behavior demonstrates a clear gender disparity, with males dominating consumption of traditional, locally accessible hand-rolled tobacco (Yasen or Bai-Jak). A reversed socioeconomic gradient was seen, with higher household income associated with increased odds of active smoking, demonstrating that financial competence directly maintains tobacco intake in this group. Furthermore, geographic context has a significant impact on clinical outcomes; while long-term conditions such as chronic cough and COPD are consistent across regions, patients in tobacco-producing areas have a significantly higher burden of acute dyspnea (28.7% vs. 21.0%) due to increased smoking severity and exposure to agricultural dusts. From a health-care system standpoint, the high prevalence of COPD among smokers (28.8%) combined with severe healthcare utilization—where more than half require four or more annual hospital admissions—places an unsustainable burden on public infrastructure. These findings show that universal national tobacco control measures do not address the distinct vulnerabilities of agriculturally specialized regions. To reduce avoidable service congestion and improve healthcare system responsiveness, public health officials must employ geographically tailored interventions. Localized respiratory screening programs, stronger regulation of unmanufactured traditional tobacco products, and community-level cessation activities tailored to the economic and environmental constraints of tobacco-producing areas are also essential methods. Limitations of the study The cross-sectional design is a methodological limitation that limits the development of causal links between tobacco production locations and smoking behavior. Furthermore, the exclusion of terminally sick or cognitively impaired patients, as well as the significant gender imbalance (74.3% female), may restrict the findings’ applicability to Thailand’s larger, mostly male smoking community. The study used self-reported data and self-rated health assessments rather than clinical diagnostics such as spirometry. This increases the possibility of recollection bias and social desirability bias, in which patients may under-report their smoking behaviors owing to the professional environment.