Modern lifestyle risks associated with alcohol consumption and cigarette smoking in Ukraine

Abstract

Background:

Substance uses confer huge risks for public health in modern society. This study aimed to evaluate current factors related to alcohol drinking and tobacco smoking in the republic of Ukraine.

Methods:

We distributed a questionnaire to healthy volunteers in four Ukrainian regions and collected 929 responses regarding demographic information, reasons for starting substance use, substance use family history, environmental factors, use pattern (internet, drinking or smoking), consequences of drinking, and insomnia. Linear regression and logistic regression analyses were performed to identify factors related to alcohol drinking and cigarette smoking.

Results:

Aging appeared to attenuate either drinking or smoking. To reduce school- or job-related stress, staying awake, peer pressure, friend-related issues, or to increase sexual desire and single parenting all would trigger drinking; male gender, family divorce, unhealthy diet and lack of awareness about harmful consequences were the main factors leading to smoking. Logistic regression suggested that education background, sleep problems, celebration events and lengthy internet work all could contribute to drinking.

Conclusions:

Various real-world factors related to substance uses were identified for the public of Ukraine. A validated instrument would help to identify risk factors in patients with substance use disorders.

Background

Drinking and smoking are the leading causes of chronic diseases worldwide (Bauer et al., 2014; Benziger et al., 2016; Collaborators, 2018; Roerecke et al., 2017) and confer huge risks for ongoing COVID-19 pandemic (“Mapping the human genetic architecture of COVID-19,” 2021; Wang et al., 2021), but the complex etiologies of substance use disorders (SUDs) remain elusive. An established view is that SUDs are caused by both environmental and genetic risks (Ajonijebu et al., 2017; Dick et al., 2018; Heath et al., 2016), which implies a possibility for dynamic etiologies as both the environment and genetic make-ups are changing constantly. Such information urges us to monitor risk factors for SUDs in the changing and modern society.

SUDs are widespread in Ukraine but the associated risks have not been investigated yet beyond depression, marital status and occupation (CMHMDA, 2020; Polshkova et al., 2016; Webb et al., 2005). According to the WHO Global Information System on Alcohol and Health, the total recorded and unrecorded alcohol per capita consumption among adults (15+) was 8.6 (in liters of pure alcohol) in 2018 (WHO, 2018), associated with increased risk of suffering adverse physical, psychological, social health outcomes and mortality (GBD 2016 Alcohol Collaborators, 2018). As well, results from Global Adult Tobacco Survey (GATS) showed that 23.0% of adults were currently smoking tobacco in Ukraine (GATS Report Ukraine, 2017), associated with increased risks of a variety of cancers, chronic obstructive pulmonary, cardiovascular, oral diseases and adverse reproductive outcomes (Savitz et al., 2006). Little is known about the risk factors for SUDs so that this epidemic urges us to interrogate factors related to substance uses in the public of Ukraine.

Currently used instrument for drinking risk assessments is the Alcohol Use Disorders Identification Test (AUDIT) which was developed 30 years ago (Babor et al., 1989) and previously used in Ukrainian adolescents (Linskiy et al., 2012). However, during the last decades, many new factors have been recognized contributing to the etiologies of SUDs. For instance, use of e-cigarettes is becoming a new form of using nicotine among adolescents (Eun et al., 2018; Hoerr et al., 2017; Juel et al., 2017; Lin et al., 2018; Weaver et al., 2018). Internet has also influenced the daily life of almost everyone (Chebli et al., 2016; L et al., 2016) and recent social media use has been associated with drinking (Brunborg et al., 2017). Reports identified that potential risk factors for addiction to the internet were: drinking behavior, family dissatisfaction, and experience of recent stressful events (Lam et al., 2009). Another study found that the risk of internet addiction was associated with cigarette smoking, alcohol drinking, drug addiction, and sexual intercourse experience (Sung et al., 2013).

To meet the modern lifestyle, we have developed an informal instrument, which considered socioeconomic status, use patterns and insomnia. The use of this instrument allowed us to collect information on factors associated with substance uses in Ukraine. Here we report the initial results from the public of Ukraine with this new instrument.

Methods

Study design

We first prepared a substance use risk factor assessment (SURFA) instrument, based on modern lifestyle and literature. The development of SURFA questionnaire consists of 6 stages: literature review, initial ideas for questionnaire content, discussions, questionnaire structure, pilot study and face validity, and questionnaire validation. After completion of the preparation, this instrument was distributed in four settlements belonging to three regions of Ukraine, those were: three million-population cities (Kyiv, Odessa, Kharkiv) and small town in the Kharkiv region (Valki). Participants were recruited between December 2016 and April 2020 in Ukraine. This research was approved by the State Institution “Institute of Neurology, Psychiatry and Narcology of the National Academy of Medical Science of Ukraine” Review Board (protocol No. 10, 13.10.2016).

Instrument preparation

An informal questionnaire was prepared covering seven categories: 1) Demographic information; 2) Reasons for starting drinking or smoking; 3) Substance use family history; 4) Environmental factors; 5) Substance (along with internet) use pattern; 6) Consequences of drinking; and 7) Insomnia. Details of this informal instrument are provided in supplemental with a understanding that more work would be required to validate it.

Response collection

The survey was carried out at the outpatient clinic of the Kharkiv National Medical University (Kharkiv), at the outpatient clinic of the Institute of Neurology, Psychiatry and Narcology of the National Academy of Medical Sciences of Ukraine (Kharkiv), at the outpatient clinic of the Main Military Clinical Hospital of the Ministry of Defense of Ukraine (Kyiv) and also at outpatient clinics in Odessa and Valki.

Surveyed contingent consisted of employees from various enterprises of settlements included in the study, who came to the indicated outpatient clinics to undergo an annual preventive medical examination or for an initial medical examination upon admission. Respondents from this contingent were selected for survey at random.

Survey completion each took approximately 45 minutes and was either self-administered or facilitated by interview with pencil-and-paper in pilot study during a face validity stage.

Participants

Nine hundred and thirty individuals took part in this study, of whom 4.2% (n = 39) were recruited from Kiev, 2.4% (n = 22) from Odessa, 91.4% (n = 851) from Kharkiv and 1.9% (n = 18) from Valki (Kharkiv region). One of the participants was excluded from analysis, as he was a 17 years old minor. Inclusion criteria were as follows: 1) 18–80 years old; and 2) ability to provide an informed consent to participate. The main exclusion criteria were: current diagnosis with psychosis, bipolar affective disorder, schizophrenia, seizures, dementia and acute suicidality who required hospitalization. The clinicians identified potential participants who were willing to participate and recruited them to this study.

Statistical analysis

Statistical analyses were performed using the SPSS version 21.0 software. Categorical data were presented as frequency and percentage. Continuous data were presented as mean ± standard deviation (SD). Chi-square tests or independent T tests were used to compare the differences in demographic characteristics between males and females. Chi-square tests were performed to analyze differences in percentage of drinking or smoking among different age groups.

In subjects who drank in the last week, linear regression analyses were performed to find out what variables (e.g., demographic, incentive, family factors, or environment) could predict drinking frequency. The same linear regression analyses were performed in those who smoke in the last week to identify variables related to smoking frequency.

In those who had drunk, logistic regression was performed to estimate the association between related variables (e.g., demographic, reasons of drinking, family factors, frequency of drinking, or frequency of smoking) and drinking consequences, with “no any medical conditions” serving as the reference level for the outcome.

P<0.05 was defined as a statistical significance; NS, for non-significant.

Results

Demographics

Participants ranged in age from 18 to 77 years, with a mean (± SD) age of 41.98 ± 12.60 years. The mean height and weight were 174.26 cm ± 10.05 cm and 81.70 kg ± 35.18 kg respectively. Participants were predominantly males (80.62%) and Ukrainians 83.0%. A mean 14.02 years of education was noticed and 80.3% of them were employed. Approximately one-half were married (56.9%). 21.9% of them had no financial difficulty and 4.6% had greatest financial difficulty. Average age starting drinking or smoking was before 19 years and average years of substance uses was more than 22 so that both alcohol drinking and cigarette smoking had the similar starting age and experience years in this cohort (Table 1).

Table 1.

Characteristics of the full sample

Variables	Full sample (n=929)	Male (n=749)	Female (n=180)	χ2/t	p
Age (M±SD)	41.98±12.60	41.46±12.37	44.13±13.36	−2.56	<0.05
BMI (M±SD)	26.33±4.84	26.40±4.61	26.03±5.71	0.81	NS
Ethnicity (valid n, %)				10.86	<0.05
Ukrainian	771 (83.44)	635 (85.12)	136 (76.40)
Russian	132 (14.29)	93 (12.47)	39 (21.91)
Other	21 (2.27)	18 (2.41)	3 (1.69)
Education level (valid n, %)				16.14	<0.01
Initial	2 (0.22)	1 (0.13)	1 (0.57)
Incomplete middle	23 (2.50)	21 (2.82)	2 (1.14)
Average	416 (45.17)	354 (47.52)	62 (35.23)
Incomplete higher	78 (8.47)	54 (7.25)	24 (13.64)
Higher	402 (43.65)	315 (42.28)	87 (49.43)
Employment (valid n, %)				2.53	NS
Yes	796 (86.52)	639 (85.66)	157 (90.23)
No	124 (13.48)	107 (14.34)	17 (9.77)
Financial difficulty (valid n, %)				15.25	<0.01
0 (enough)	203(22.51)	164 (22.62)	39 (22.03)
1	146 (16.19)	132 (18.21)	14 (7.91)
2	230 (25.50)	186 (25.66)	44 (24.86)
3	213 (23.61)	158 (21.79)	55 (31.07)
4	67 (7.43)	52 (7.17)	15 (8.47)
5 (hardly enough for food)	43 (4.77)	33 (4.55)	10 (5.65)
Live with (valid n, %)				0.56	NS
Live with both parents	760 (82.25)	611 (81.79)	149 (84.18)
Live with one parent	164 (17.75)	136 (18.21)	28 (15.82)
Family in jail (valid n, %)				0.22	NS
Yes	8 (0.87)	7 (0.94)	1 (0.57)
No	916 (99.13)	741 (99.06)	175 (99.43)
Family divorce (valid n, %)				2.20	NS
Yes	263 (28.59)	205 (27.52)	58 (33.14)
No	657 (71.41)	540 (72.48)	117 (66.86)
Marital status (valid n, %)				0.045	NS
Never married	202 (25.28)	168 (25.69)	34 (23.45)
Married	528 (66.08)	443 (67.74)	85 (58.62)
Divorced	69 (8.64)	43 (6.57)	26 (17.93)
Age began alcohol (M±SD, n)	18.91±3.17 (822)	18.95±3.21 (671)	18.71±2.97 (151)	0.76	NS
Drinking years (M±SD, n)	22.67±11.80 (822)	22.42±11.69 (671)	23.68±12.17 (151)	−1.32	NS
Smoking (valid n,%)				31.21	< 0.001
No	552 (59.42)	412 (55.01)	140 (77.78)
Yes	377 (40.58)	337 (44.99)	40 (22.22)
Age began smoking (M±SD, n)	18.30±3.85 (477)	18.12±3.74 (435)	20.19±4.44 (42)	−3.37	<0.01
Smoking years (M±SD, n)	22.86±11.76 (477)	22.11±11.66 (435)	20.19±12.58 (42)	1.54	NS

Notes: Continuous variable expressed as mean ± SD; Categorical data were presented as frequency and percentage; BMI, Body mass index

Age effect

Majority of the participants (88.48%) had drinking experience and 51.35% had smoking experience. The participants were divided into three age groups, younger (18–44), middle (45–59) and older (above 60). Overall, prevalence of either drinking or smoking decreased with aging in a statistically significant manner. This reciprocal correlation remained statistically significant for drinking in females (Table 2). Of the note, aging had little effect on male drinking but statistically marginal effect on male smoking. The female smoking sub-cohort was too small to evaluate aging effects.

Table 2.

Percentage of drinking and smoking in different age groups

Age range (years old)	Drinking						Smoking
	Full sample (n=822)	Prevalence (%)	Male (n=671)	Prevalence (%)	Female (n=151)	Prevalence (%)	Full sample (n=477)	Prevalence (%)	Male (n=435)	Prevalence (%)	Female (n=42)	Prevalence (%)
18–44	480	91.1	402	90.7	78	92.9	290	55.0	263	59.4	27	32.1
45–59	285	86.6	222	88.4	63	80.8	160	48.6	148	59.0	12	15.4
Above 60	57	78.1	47	85.5	10	55.6	27	37.0	24	43.6	3	16.7
χ2	12.35		1.99		16.25		9.86		5.09		6.85
p	<0.01		NS		<0.001		<0.01		NS		NS

Daily use pattern

We asked about daily use of a substance from Monday to Sunday. 534 (57.48%) of them responded for drinking and 394 (42.41%) responded for smoking. Nobody drank on Monday, a half of participants drank on Friday and all of them drank on Saturday, uncovering a weekend drinking pattern (Figure 1a). Smoking pattern was different as there was no day-to-day effect so that the cigarette consumption levels remained the same throughout the week (Figure 1b). No age effect on the day-to day drinking/smoking patterns was found.

Figure 1.

a. Pattern of alcohol drinking (standard drinks) among those who drank in the last week (n=534). b. Pattern of tobacco smoking (cigarette number) among those who smoked in the last week (n=394).

Factors associated with drinking

In those who drank in the last week (valid n=506), we identified eight factors correlated with drinking frequency by linear regression with all demographic, incentive, family or environment factor in the whole model. The top three were sleep deficiency, friend-related issues and school- or job-related issues (p<0.001). Other factors were “too much to do”, “to reduce tension”, “to increase sexual desire”, “single parent with a kid”, and “peer pressure” (Table 3).

Table 3.

Linear regression analysis for prediction of drinking frequency in whoever drank in the last week

Predictor variables	β	95%CI	p-value
Reduce tension	−0.13	−0.55 – −0.12	<0.01	F = 18.211
Increase sexual desire	−0.13	−1.64 – −0.39	<0.01	p < 0.001
Wake you up	0.22	0.59 – 1.45	< 0.001	r2 = 0.357
Peer pressure	−0.10	0.58 – 1.68	<0.05
Friend-related issue	−0.23	−0.08 – 1.04	< 0.001
Something bothers in school or job	0.22	0.28 – 1.43	< 0.001
Single parent with a kid	0.13	−0.06 – 0.00	<0.01
Too much things to do	−0.14	−0.04 – 1.91	<0.01

Note: β: standard regression coefficient; CI: confidence interval

Factors associated with cigarette smoking

In those who smoked in the last week (valid n=377), we identified four factors correlated with smoking frequency by linear regression with all demographic, incentive, family and environment factor in the whole model. These factors were gender (male as risk with p<0.01), family divorce (p<0.01), unhealthy diet (p<0.01) and school issue (p<0.05) (Table 4). Divorces triggered both types of substance uses (Table 4).

Table 4.

Linear regression analysis for prediction of smoking frequency in whoever smoked in the last week

	β	95%CI	p
Gender	−0.17	−55.11 – −14.91	<0.01	F = 6.502
Family divorces	0.16	6.69 – 31.75	<0.01	p < 0.001
Smoking harm teaching from school	−0.11	−28.98 – −1.18	<0.05	r2 = 0.081
Unhealthy food habit	0.14	0.54 – 3.03	<0.01

Note: β: standard regression coefficient; CI: confidence interval.

Variables related to drinking consequences

By logistic regression among those who drank (valid n=457, without missing data), we revealed seven factors that could predict drinking consequences (1=with medical problems, 0=without medical problems) as listed in Table 5. Among them, peer pressure was the most risky factor (OR=2.07) of medical problems. Education level, celebration, sleep problem, drinking frequency, “internet for work” and interpersonal relationship problem also contributed to the dinking consequences.

Table 5.

Logistic Regression analysis for prediction of medical conditions in those who drank

	OR	95%CI	p
Education level	0.65	0.47 – 0.90	<0.05
Drinking for celebrating success	1.25	1.02 – 1.52	<0.05
Drinking for Peer pressure	2.07	1.01 – 4.22	<0.05
Drinking for Interpersonal relation problem	0.33	0.12 – 0.91	<0.05
Time of internet for work	1.04	1.01 – 1.07	<0.05
Sleep problems	1.18	1.05 – 1.33	<0.01
Drinking frequency	1.12	1.01 – 1.25	<0.05

Note: OR= odds ratio; CI= confidence interval.

Discussion

We initiated a systemic survey on factors related to drinking and smoking and obtained novel findings based on Ukrainian cohorts.

This is the first time we interrogate the environmental factors those trigger substance uses and misuses in the public of Ukraine. By comparison, there seem to be different reported reasons for drinking vs smoking. Drinking helped “relief” whereas smoking was to deal with interpersonal issues. More interestingly, unhealthy diet may contribute to smoking behavior, which has been suggested before in a Brazilian adult cohort (Francisco et al., 2019). Aging attenuates the tendency to smoke but not drink. Peer pressure remains an independent factor for substance misuses, as previously recognized (Morris et al., 2020; Studer et al., 2014).

Substance use behaviors have strong genetic components so that Ukrainians are representative of Europeans genetically and our current findings could be extrapolated to other European ancestry (Bashynska et al., 2019).

Limitations and future directions

Although this study adds to the research fields by using a novel questionnaire, there are major limitations. First, we need more female participants because the current cohort had only 180 women so that we have less confidence on findings in the females. A second limitation is that we were not targeting patients with SUDs so that many questions got no valid scores and we got limited disease-related information including family history and comorbidity. Third, participants were recruited mainly from Kharkiv, a city in eastern Ukraine. We would need participants also from Western and Northern parts of the country, to increase generalizability of the findings. Fourth, electronic devices such as use of e-cigarettes should be considered as well in the questions for cultures to which e-cigarettes are available. Future studies should validate the instrument among patients with SUDs, providing insights into etiologies of SUDs and also facilitating the development of precision healthcare.

In conclusion, as a public health factor, lifestyle is associated with substance uses and misuses in our modern society, including family health, personal relationship and diet among other factors.