High frequency of chronic cough and sputum production with lowered exercise capacity in young smokers

Abstract

Objectives. The aim was to evaluate how cigarette smoking is associated with respiratory symptoms, fitness, and anthropometric measures in young smokers. Methods. The prevalence of smoking was investigated in a cohort of young military draftees (n = 1130; 98% between 18–21 years of age) in Northern Finland. The associations of smoking with respiratory symptoms, physical fitness (12-min running test), education, and anthropometric measures were analysed using a self-reported questionnaire with high response rate (80%). Results. Almost half (46.5%) of the young males were daily smokers, 17.4% being occasional smokers. The prevalence of self-reported chronic cough and sputum production was high in daily smokers (40.7%) and occasional smokers (26.9%) compared to non-smokers (12%). These symptoms were significantly associated with the smoking history. Aerobic fitness was worse in regular smokers compared to non-smokers (P < 0.001). Smokers had a higher body mass index than non-smokers (P = 0.035). In the regular smokers, the more active the subjects were in sports, the less they smoked when evaluated by pack year history (P < 0.001). Smokers had a lower educational level than occasional smokers or, especially, non-smokers (P < 0.001). Conclusions. The frequency of young smokers with chronic cough and sputum production was very high, posing a serious risk to their future health.

Key words::

• Chronic cough and sputum production
• fitness
• smoking
• young adults

Abbreviations
BMI	=	body mass index
COPD	=	chronic obstructive pulmonary disease
cm	=	centimetre(s)
ETS	=	environmental tobacco smoke
GOLD	=	The Global Initiative for Chronic Obstructive Lung Disease
kg	=	kilogram(s)
K-W	=	test Kruskal-Wallis test
max	=	maximum
min	=	minimum
r(s)	=	Spearman's correlation coefficient
VIF	=	variance inflation factors

Key messages

• Surprisingly many young men (> 45%) in Northern Finland are daily smokers, reflecting great regional differences within the country.

• Chronic cough and sputum production, potential risk factors for COPD development, are very common already in young cigarette smokers.

• Smoking impairs physical fitness in young smokers and is associated with lower educational level and a passive life-style, i.e. physical inactivity, and elevated body mass index.

Introduction

Cigarette smoking is still common, especially among adolescents and young adults. Based on recent data from Tobacco Atlas, 30%–39.9% of males use tobacco in many European countries including Finland (1). In Finland 29% of young adults smoked in 2007 (2). Smoking in turn correlates with the development of chronic obstructive pulmonary disease (COPD) (3,4). Young adults in their twenties or thirties may have a smoking history of 10–15 years, which is long enough for the development of COPD. Since COPD is predominantly a progressive disease, efforts need to be focused on preventing the initiation of smoking or alternatively motivating to give up smoking as soon as possible. However, relatively little is known about the smoking habits of young smokers, how smoking actually impacts on airway symptoms in young smokers, and whether smoking can contribute to exercise capacity and fitness, or to parameters such as height and weight in young smokers. This study was undertaken to investigate especially these issues in young cigarette smokers.

Recent studies have revealed that chronic cough and/or sputum production without airway obstruction (Stage 0 in the previous Global Initiative for Chronic Obstructive Lung Disease (GOLD) classification of COPD (4)) is common (10%–40%) among smokers and related to the risk of COPD development and increased mortality (5–7). However, in those studies, the populations have included large numbers of middle-aged smokers, and the analyses have not focused on younger subjects.

Smoking and physical fitness have been mostly studied with military personnel (8–13). The physical fitness tests conducted by the military are a cost-effective, rather easily available and objective method for estimation of physical fitness capacity. The results of these and other studies (14–16) on this issue including young adults have concluded, not surprisingly, that smoking negatively correlates with physical fitness. In these studies, the sample sizes have been variable, especially the smoking subgroups have often been small, and smoking history not properly described; i.e. no information about the intensity or years of smoking in a study cohort has been available. The age of the population has also varied widely, and often different age groups have not been analysed separately. Therefore, the effects of smoking on the physical fitness in young adults in their twenties are still unresolved.

Smoking has been claimed to evoke weight reduction (1,17–21) and smoking cessation to cause weight gain in adult populations (22). The metabolic effects of smoking on weight have been studied to unravel the underlying mechanisms, which appear to be complicated (23,24). There have been conflicting results from studies on young populations reflecting the difficulty in controlling the plethora of confounding variables. In addition, smoking exposure may be insufficient to detect the weight-attenuating effect of tobacco in a young population. Moreover, experimental studies have indicated that nicotine accumulation and distribution in the brain is age-dependent, which in turn may have an influence on both the nicotine dependence and physical variables of young smokers (25). It is, however, unknown whether active smoking has any effect on the height gained during adolescence.

This study estimated the overall prevalence of smoking in a representative population of Finnish young male adults living in the northern parts of Finland where smoking is more prevalent than in the urban areas of Southern Finland or in Northern Sweden (2,26,27). The specific aim was to find out whether the harmful effects of smoking both on the respiratory symptoms and physical fitness are detectable already in the age group of young adults who have a relatively short smoking history. Additionally smoking status was correlated with education and anthropometric measures such as height, weight, and body mass index (BMI).

Methods

Study design and data collection

This survey was conducted in Northern Finland during autumn 2008 to spring 2009. The study population consisted of military recruits from Lapland Antiaircraft Regiment and Light Infantry Brigade, part of the Northern Command of the Finnish Defence Forces. About 40% of these draftees come from Lapland and the rest (60%) from Northern Ostrobothnia, the two most northern regions of Finland. In Lapland, approximately 81%–82% of an age group of male adolescents is recruited for military service each year. Military service is voluntary for women in Finland. All draftees were asked to fill in a specific questionnaire about their sex, age, height, weight, education, physical fitness, detailed smoking history, daily exposure to environmental tobacco smoke (ETS), and symptoms of cough and sputum production. The questions about the smoking habits were based on a Finnish nation-wide study questionnaire, the Adolescent Health and Lifestyle Survey (2). The study was approved by the Ethical Committee of the Lapland Central Hospital in Rovaniemi, Finland.

Measurements

Respiratory symptoms indicative of self-reported chronic bronchitis were assessed by the question ‘Have you had cough with sputum production on most days or nights for at least 3 months yearly?’ as these symptoms predict COPD development (4,6,26,27). A Cooper's test result, a measure used in this study, was self-reported in the questionnaire. Cooper's test is a 12-minute running test for physical fitness that all conscripts are required to undertake during their first 3 weeks of service if there is no health reason for exemption. Subjects run as far as possible within 12 minutes around a running track. The recruits were encouraged to do their best since they could obtain extra leave when the test result exceeded 3000 metres. Alternatively, a poor result meant that the recruit had to participate in an extra fitness-enhancing programme, something they were keen to avoid. The running distance was recorded by the army personnel.

The subjects were classified as regular smokers, occasional smokers, non-smokers, and ex-smokers. A regular smoker was defined as a person who currently smoked at least one cigarette daily. An occasional smoker was defined as a subject who currently smoked less than one cigarette daily or reported having smoked >50 cigarettes in the past, and at the time considered himself as a smoker. A non-smoker was a person who had smoked <50 times in his life and did not consider himself as a smoker. Ex-smokers were subjects who reported that they had stopped smoking. Ever smokers reported smoking initiation age, daily tobacco usage, and number of years smoked. Here grams of tobacco smoked daily consist of cigars (3 g/cigar) and manufactured or self-rolled cigarettes (1 g/cigarette). One pack year was defined as 20 g of tobacco smoked per day for 1 year. None of the draftees smoked pipe tobacco.

Both body weight and height were self-reported in the questionnaire. Body mass index is the body weight (kg) divided by the square of height (m). Physical activity level was assessed with a question ‘How would you describe yourself as a sportsman or sportswoman?’ The options were ‘I compete in some sport’, ‘I'm very active in sports, but I don't compete’, ‘I'm a “couch potato”—I only do sports if I have to’, and ‘Something else, what?’ Subjects were divided into subgroups ‘Active’, ‘Moderate’, and ‘Passive’. Subjects who competed or were otherwise active in sports were considered as ‘Active’. Based on specific descriptions given by subjects, ‘Moderate’ were subjects who answered ‘Something else’. “Couch potatoes” were considered as ‘Passive’.

For the linear and multinomial logistic regression analyses, education was categorized into two groups. The lower education group included subjects with primary school or vocational school. The higher education group had graduated from high school or had an academic degree.

Study subjects

The study originally consisted of a total of 1186 participants. All women were excluded from the data because of the small number of female participants (n = 12). From the 1174 male participants, 11 subjects with unreliable or incomplete filling of the questionnaire were also excluded. All ex-smokers (n = 33) were excluded from analysis because of the small number in this subgroup. Thus, the final study population consisted of a total number of 1130 male subjects. Mean age of the whole study population was 19.4 years (min 18, max 25), 98% being between 18 and 21 years.

Statistical analysis

The data analysis was performed using SPSS for Windows 16.0 software. Relationships between smoking status and categorical variables were analysed using cross-tabulation and chi-square tests. Kruskal-Wallis test (K-W test) was used to evaluate the relationship between categorical variables and variables of tobacco usage (daily usage, smoking years, and pack years), because all three variables had a right-skewed distribution. Multinomial logistic regression was used to evaluate the relationship between self-reported symptoms of prolonged cough and sputum production, smoking habits, and potential confounders (BMI, age, education, and physical activity). In the Cooper's test result, the differences between the smoking groups were tested first with one-way ANOVA. If the difference between the groups reached significance, the groups were further analysed with Tukey's post-hoc test for pairwise comparisons. Linear regression was used to evaluate the relationship between Cooper's test result and potential covariates. We evaluated the potential collinearity between explanatory variables using variance inflation factors (VIF) (28). Scatter plot diagrams and Spearman's correlation coefficient (r(s)) were used to illustrate the correlation of the Cooper's test result to daily tobacco usage, smoking years, and pack years.

One-way ANOVA and an additional Tukey's test were also used to evaluate differences in age and physical measures in the smoking groups. Scatter plot diagrams and Spearman's correlation coefficient (r(s)) revealed the correlation between BMI with the variables of daily tobacco usage.

Results

Smoking habits and self-reported symptoms of prolonged cough and sputum production

Of the 1130 subjects, 525 (46.5%) were regular (daily) smokers, 197 (17.4%) occasional smokers, and 408 (36.1%) non-smokers. The median pack year was 3.6, but the variability was extensive from 0.04 even to 29 (Table I). Approximately one in ten had already smoked at least 10 pack years. The mean age of starting to smoke was 15 years (min 8, max 23).

Table I. Cohort's smoking history in pack years.

Pack years	n(%)
0–2.99	189 (37.9)
3–5.99	163 (32.7)
≥ 6	147 (29.5)
Total	499 (100.0)

Symptoms were frequent and significantly more common in regular smokers than in occasional smokers, and more common in occasional smokers than non-smokers (Figure 1: 40.7% versus 26.9% versus 12%, P < 0.001). The difference between smoking subgroups remained strongly significant (P ≤ 0.001) after taking into account the potential confounders (BMI, age, education, physical activity) in multinomial logistic regression analysis. The median of pack years was 4.5 with symptoms, 3.0 with no symptoms, and 3.85 for those responding ‘I don't know’ (K-W test P < 0.001). The median of daily tobacco grams was 18 with symptoms, 15 with no symptoms, and 16 for those responding ‘I don't know’ (Kruskal-Wallis test P value < 0.001). Additionally, the median smoking years was 5.0 with symptoms, 4.0 with no symptoms, and 4.0 for those responding ‘I don't know’ (K-W test P value = 0.007).

Figure 1. Answers to the question ‘Have you had cough with sputum production on most days or nights for at least 3 months yearly?’ according to smoking status. Self-reported symptoms of prolonged cough and sputum production were very common in young smokers, and more common in smokers than in non-smokers. Chi-square test P value < 0.001.

Among non-smokers, 22.4% (n = 91) had had daily exposure to tobacco smoke. Of these exposed non-smokers, only six had 1–5 hours of exposure daily, and none had more than 5 hours of exposure. Their symptoms did not differ significantly from non-exposed non-smokers.

Smoking habits and physical fitness

The performance of regular smokers in the 12-min running test was worse than their non-smoking counterparts (Table II, P < 0.001). Occasional smokers in turn had a significantly better mean test result than regular smokers (P < 0.001). The mean results of occasional smokers and non-smokers did not differ significantly.

Table II. Subject characteristics and means of Cooper's test result.

	Non-smoker		Occasional smoker		Regular smoker		Total
	n	mean (SD)	n	mean (SD)	n	mean (SD)	n	mean (SD)	P value of F test
Age (year)	406	19.4 (0.8)	197	19.4 (0.6)	525	19.5 (0.9)	1128	19.4 (0.8)	>0.05
Height (cm)	408	178.0 (6.3)	196	178.3 (6.4)	524	177.9 (6.8)	1128	178.0 (6.5)	>0.05
Weight (kg)	408	73.4 (9.6)	197	74.7 (9.8)	523	74.7 (10.3)	1128	74.2 (9.9)	>0.05
BMI (kg/m^2)	408	23.1 (2.5)	196	23.5 (2.6)	523	23.6 (2.7)	1127	23.4 (2.6)	0.042^a
Cooper's test result (m)	396	2537.1 (355.4)	188	2510.7 (325.5)	508	2376.6 (405.0)	1092	2457.9 (382.0)	<0.001^b,c

^aRegular smokers compared to non-smokers Tukey's test P value 0.035.

^bRegular smokers compared to non-smokers Tukey's test P value < 0.001.

^cRegular smokers compared to occasional smokers Tukey's test P value < 0.001.

The linear regression analysis revealed that smoking and BMI were negatively correlated with the distance covered in the running test (Table III). As expected, the more active the subjects were in sports, the better was the test result. Education also had strong positive association with test performance—the result was better in subjects who had graduated from high school or had an academic degree than in those who had graduated from vocational school or only primary school. The strongest influence was noted with the physical activity, the next being education, and the third with smoking and BMI, with the last two factors having similar effect. The magnitude of collinearity between explanatory variables was not high as estimated by the VIF statistic.

Table III. Outcome of linear regression analysis.^a

	Unstandardized coefficients			Collinearity statistics
	B	SE	P value	VIF
(Constant)	2975.2	244.6	<0.001
Smoking	−33.1	11.6	0.004	1.2
Activity in sports	198.0	11.9	< 0.001	1.1
BMI	−37.0	3.7	<0.001	1.0
Education	70.5	21.7	0.001	1.2
Age	7.1	12.2	0.56	1.0

^aDependent variable: Cooper's test result (m).

The mean distance covered in the 12-minute running test did not correlate to daily grams of tobacco or smoking years, but there was a weak negative relation to the pack years (r(s) = –0.113, P = 0.013) i.e. the more pack years in their smoking history, the shorter distance they ran in the Cooper's test.

Anthropometric measures

The mean age of the three smoking groups did not differ significantly (Table II). Neither mean height nor mean body weight differed significantly between the smoking subgroups, but the BMI was significantly greater (P = 0.035) in regular smokers than non-smokers. Instead, the BMI of non-smokers and occasional smokers did not differ significantly from each other. BMI did not correlate with daily tobacco usage or calendar years of smoking (no visual correlation and Spearman's coefficient insignificant), but BMI had a weak positive correlation with pack years (r(s) = 0.103, P = 0.022).

Physical activity

The amount of physical activity differed between the different subgroups (Figure 2; chi-square test P < 0.001). Regular smokers were clearly less active than non-smokers. Among regular smokers, the median of pack years was 2.78 with subjects defined as ‘Active’, 3.6 for ‘Moderate’, and 4.50 for ‘Passive’. The median of daily amount of tobacco was 14.0 for ‘Active’, 16.0 for ‘Moderate’, and 20.0 for ‘Passive’. The median of years of smoking was 4.0 for ‘Active’, 4.0 for ‘Moderate’, and 5.0 for ‘Passive’ (K-W test P value for all three variables < 0.001). Therefore, the more active the subjects were, the less they smoked daily and the less they had smoking years and pack years in history.

Figure 2. Activity in sports according to smoking status. Regular smokers were most frequently passive in sports, non-smokers being the most active. Chi-square test P value < 0.001.

Education

Smokers had less commonly graduated from high school or had an academic degree and more frequently completed vocational school or only primary school than had non-smokers (Figure 3; chi-square test P value < 0.001). The difference was more remarkable if non-smokers were compared to regular smokers than if they were compared to occasional smokers. Few of the respondents (n = 11; 1%) had an academic degree.

Figure 3. Educational level according to smoking status. Smokers had lower educational level than non-smokers. Also occasional smokers were more educated than regular smokers. Chi-square test P value < 0.001.

Discussion

Surprisingly many young men (> 45%) coming from northern regions of Finland are daily smokers; if also occasional smokers are included the number is > 60%. This study revealed that symptoms of self-reported chronic bronchitis (i.e. chronic cough and sputum production on most days for at least 3 months yearly) were very common, over 40% in regular smokers. This is important since the mean age of these adolescents was 19 years. The more a subject smoked, the more likely he would display the symptoms. In young men, smoking is strongly associated with lowered aerobic fitness as assessed by the 12-min running test even when physical activity, education, and BMI are controlled. Based on this study, regular smokers had higher BMI than non-smokers. Smoking was also related to lower physical activity.

Despite the decreasing number of smokers in Finland, there are approximately 1 million smokers in a population of 5 million. The average age to start smoking in Finland is 14 years of age. In our study, the mean age when young males started to smoke was 15 years, and the number of daily smokers was relatively high (46.5%) possibly reflecting the major variability in smoking habits in different parts of the country. There are no other studies focusing on the smoking habits of young males in Northern Finland, but it is known that almost ten years earlier smoking was relatively common in the adult population (32%, 46 years of age) of this area (26), the corresponding number from Northern Sweden being 26% (27). In our study we used the same questions for smoking habits as the previous nation-wide Adolescent Health and Lifestyle Survey (2). Although we have some minor differences in the definitions for smokers, the number of daily smokers is the same as it would be according to their definitions. Therefore, smoking in this cohort of Northern Finland is comparable to that recent study. Based on it 29% of 18-year-old adolescents smoked in Finland in 2007. The cut-off for ever smoking in any studies conducted in adult populations is 100 instead of 50 cigarettes, which was used in our study and some other studies with adolescents in Finland (2,29). This difference in the definition needs to be taken into consideration when the prevalence of smoking is being compared. Our result highlights that the specific geographical characteristics, both in youngsters and adults, are very important in the planning of the smoking cessation counselling and education throughout the country.

A remarkable number of ‘healthy smokers’ are known to suffer from chronic cough and mucus production. A recent large Finnish adult twin study revealed that the risk of chronic bronchitis increased about 1.5-fold by each amount category of daily cigarettes, but also that not only moderate and heavy, but also former and light smokers had significant risk for chronic bronchitis (30). Most previous studies have included smokers of different ages and generally older smokers than in our study. In Copenhagen City Heart Study, for example, the prevalence of chronic bronchitis was 23.4% in male heavy smokers 65 years of age or older (31), and in an Italian study the corresponding symptoms could be seen almost in 40% of adult smokers (32). Young age, however, does not protect from chronic bronchitis or COPD (6,33). One study has described a dose-response relationship between smoking in adolescents and already lowered FEV1/FVC and FEF25-75 as evidence of early airway obstruction and small airway disease (34). The recent study of Guerra et al. (7) found that chronic cough and sputum production represent early markers for future COPD development and increased mortality but only among the group of younger smokers, i.e. in that particular study less than 50 years old. Those findings are in line with another large study (5) in which 12% of young and middle-aged (over 5000 individuals, aged between 20 and 44 years) smokers suffered from chronic cough and/or phlegm without airway obstruction. A considerable number of young people already suffered from COPD (GOLD stage 1+, 3.6%), and the presence of chronic cough/phlegm also significantly predicted COPD development (6). The symptoms of chronic cough and sputum production in the present study were much more common than in many other studies especially if the relatively short smoking history is taken into consideration. The prevalence of symptoms was relatively common also in non-smokers (12%) which could not be explained by ETS, because the prevalence of symptoms in exposed non-smokers did not differ significantly from non-exposed non-smokers.

High numbers of symptoms in our study may be related to several reasons, one of those being the recognition of the existent symptoms. Our recent study, which was originally focused on symptom-free smokers, concluded that a remarkable number of those smokers reported respiratory symptoms, i.e. cough and/or sputum production, if a detailed questionnaire about the symptoms was conducted (6,35). In the present study, all subjects had been given information/lecture about the effects of smoking during their first weeks in the military service, which may have affected the final answers to the questionnaire. Since the questionnaire was anonymous, it could not lead to extra benefits in the army. Respiratory infections are frequent in the army, especially during the winter period, and could be responsible for some of the respiratory symptoms reported by both smokers and non-smokers. It can be concluded that a very high number of young daily smokers, younger than those investigated in other studies, report symptoms of chronic cough and sputum production.

Previous studies investigating the relationship between smoking and aerobic fitness involving relatively young populations have reported almost unanimous results—smoking has a deleterious effect on endurance performance (8–16). Our study confirms that this is also true in adolescence. One study on smoking in young adults (12) used Cooper's test as a physical fitness parameter and found that test performance was inversely related to daily cigarette consumption and years of smoking. Impaired performance was seen even in light smokers (1–10 cigarettes/day) with less than 2 years of smoking history. In our study, the result of Cooper's test was indeed inversely related to pack years, although the association was not strong. In a prospective twin study from Finland, physically inactive co-twins had an increased risk for smoking in early adulthood (36). In our study, smokers were also less active in sports than non-smokers, but it did not totally explain the fitness impairment in smokers. To conclude, smoking is not only associated with sedentary life-style but also poor aerobic fitness already during early adulthood.

In most adult populations, smoking is associated with lower BMI (15–19), and the difference appears to become more pronounced with increasing number of smoking years (20,21). Increasing frequency of smoking, however, does not seem to lower the BMI linearly, but there seems to be a U-shaped relation, with moderate smokers being the leanest (18,20,21). In most cross-sectional studies, adolescent smokers (< 20 years old) have had higher weights than non-smokers (37). Prospective studies seem to support the hypothesis that smoking might have a small weight-attenuating effect, but the process takes time, perhaps even decades (37–42). Many studies have reported contradictory results with gender- and/or race-dependent outcomes. Two cross-sectional studies (43,44) with large sample sizes of young adults in their twenties (over 30,000 participants, mean age 19 and 20) found a similar association between smoking and body weight, with smoking being negatively related to body weight especially in white males (approximately –1 kg). In our cohort of young males, mean BMI was higher among daily smokers than non-smokers, one possible reason being their less physically active life-style. The more that the draftees had smoked, the higher their BMI. This and other cross-sectional observations reporting a positive relationship between smoking and body weight may not simply be attributable to insufficient exposure to cigarette smoke necessary for weight loss since there are many other contributors, for example alcohol consumption or consumption of a high-fat diet as well as physical inactivity. On the other hand, smoking could hypothetically have a weight-increasing or no metabolic effect in growing adolescents—an effect which could later be reversed. Overweight adolescents could also be at a higher risk to start smoking. Smoking is recognized as a weight control strategy especially in white young females (37) and has been associated to recurrent intentional weight loss episodes in adolescents (45). Interestingly, a recent prospective twin study (29) concluded that smoking in adolescence is a risk factor for abdominal obesity in early adulthood among both genders and for overweight in young women. Confounders (physical activity, education, dietary behaviour) did not fully explain the association. More longitudinal research with comprehensive adjustment for confounding variables is needed.

Despite the number of studies investigating the relationship between smoking and BMI, very few studies have examined the relationship between active smoking and height. Studies indicate that exposure to environmental tobacco smoke during pregnancy leads not only weight decreases in the infant but also reduced birth length (46). Two out of three prospective studies have reported a small but statistically significant height reduction in young smokers (38,40). The third study failed to find significance between smoking and reduced body height (39). These follow-up studies included adolescents mainly aged 11 to 17 years of age with relatively small numbers of regular smokers. Our study evaluated a representative number of young adults who have just completed their height growth. If active smoking really had a substantial effect on height development during adolescence, then young adult smokers would be significantly shorter than non-smokers. According to our findings, smoking has no major effect on the height of young smokers. It needs to be emphasized that when evaluating the association between height and smoking, several other aspects, such as detailed genetic information including the height of the relatives and nutritional history of the participants, need to be taken into consideration, which have not been included in most previous studies or in our study.

Approximately 5%–10% of draftees have a diagnosis of asthma. One obvious limitation of our study is that we did not have exact individual information about allergies and asthma. It is worth noting, however, that during the conscription process, draftees with symptomatic asthma are quickly discovered and treated. In 2009 about 5% of the draftees in our cohort area used asthma medication daily and experienced themselves as symptom-free. Draftees with persistent symptomatic asthma are exempted from military service. Draftees with pollen allergy are enlisted to the army in the winter and complete their service mostly before the allergy season. Therefore it is unlikely that symptoms of asthma or pollen allergy impaired their performance in the fitness test. The goal of this study was to investigate smoking and its prevalence, respiratory symptoms, and anthropometric measures in young smokers. Further studies are needed to investigate asthma and allergy-related symptoms in young smoking adults.

To conclude, there is extensive variability in the smoking habits of young adults in different parts of one country, even in Europe. If both daily and occasional smokers are included, these young (19 years, mean) males displayed signs of self-reported chronic bronchitis in most of the cases, this being suggestive of a real risk for COPD development. Despite their young age, their physical fitness was already worse than non-smokers, and there was an association between smoking history, higher BMI, and sedentary life-style.

Acknowledgements

The authors thank General, Chief Medical Officer Pentti Kuronen, Finnish Defence Forces, Chief Medical Officer Markku Kerola, Northern Command, Captain Jyrki Sirkeinen, Jaeger Brigade Sodankylä, and Medical Officer Sakari Unga, Jaeger Brigade, Sodankylä.

Declaration of interest: This research was funded by a governmental subsidy from the Ministry of Social Affairs and Health, a governmental subsidy for health science research (EVO) in Rovaniemi, and Helsinki Finnish Antituberculosis Association Foundation, Yrjö Jahnsson Foundation, Sohlberg Foundation, Cancer Society of Northern Finland, Väinö and Laina Kivi Foundation.