Prevalence and determinants of SO use
Among the sample, 3.8% used SO, and 57.4% used medicaments other than SO. Among men, 3.0%, and among women 5.0% used one or more SO. There were 0.1% (6) barbiturate users, 2.5% (109) benzodiazepine users, 0.02% (1) carbamate user, 0.2% (7) users of other SHA, and 1.1% (49) opioid users. Among men, 3.0%, and among women 5.0% used SO (Table
1).
Table 1
Sedatives, hypnotics or anxiolytics (SHA), opioids, and other medicaments users
Total | 2.1 | 0.9 | 55.8 | 41.2 | 100.0 | 3.5 | 1.5 | 60.6 | 34.4 | 100.0 |
All individuals | Chi2 27.5; df 3; p < 0.001; w .08 |
All medicament users | Chi2 6.6; df 2; p <.05; w .05 |
Age (years) | | | | | | | | | | |
20–39 | 1.5 | 0.5 | 31.5 | 66.5 | 100.0 | 1.0 | 1.1 | 39.9 | 58.0 | 100.0 |
40–59 | 1.0 | 1.2 | 47.7 | 50.1 | 100.0 | 3.0 | 1.7 | 62.6 | 32.8 | 100.0 |
60–79 | 3.8 | 0.9 | 82.7 | 12.6 | 100.0 | 6.8 | 1.6 | 80.3 | 11.3 | 100.0 |
All individuals | Chi2 490.5; df 6; p < 0.001; w .46 | Chi2 367.5; df 6; p < 0.001; w.40 |
All medicament users | Chi2 7.7; df 4; ns; w .08 | Chi2 13.7; df 4; p < 0.01; w .10 |
Education (years) | | | | | | | | | | |
< 10 | 2.9 | 0.8 | 69.6 | 26.7 | 100.0 | 5.8 | 1.8 | 74.4 | 18.0 | 100.0 |
10 | 1.8 | 1.1 | 43.2 | 53.9 | 100.0 | 2.0 | 1.6 | 51.0 | 45.4 | 100.0 |
> 10 | 1.0 | 0.6 | 51.4 | 47.0 | 100.0 | 2.0 | 0.3 | 57.5 | 40.3 | 100.0 |
All individuals | Chi2 149.5; df 6; p < 0.001; w .26 | Chi2 181.1; df 6; p < 0.001; w.28 |
All medicament users | Chi2 5.4; df 4; ns; w .07 | Chi2 13.4; df 4; p < 0.01; w .09 |
SO use was not more prevalent among smokers or alcohol risk drinkers than among subjects who neither smoked nor drank alcohol in a risky way. Among male current cigarette smokers who also drank alcohol in a risky way, there were 2.3% and among female current cigarette smokers who also drank alcohol in a risky way there were 4.8% who also used SHA or opioids (Table
2). Smokers and risk drinkers were 1.6 to 3.7% among men and 4.7 to 5.1% among women. No associations were found between smoking or alcohol risk drinking and SO use: None of the smoker and risk drinker subgroups of the sample was different from subjects who had no medicament use as the reference group with respect to SO use in a multinomial logistic regression analysis after controlling for age and sex (Table
3)
Table 2
Sedatives, hypnotics or anxiolytics (SHA), opioids, and other medicaments users
Total | 2.1 | 0.9 | 55.8 | 41.2 | 100.0 | 3.5 | 1.5 | 60.6 | 34.4 | 100.0 |
Current smokers, alcohol risk drinkers | | | | | | | | | | |
Smoker, risk drinker | 1.9 | 0.4 | 37.9 | 59.8 | 100.0 | 4.8 | 0.0 | 58.7 | 36.5 | 100.0 |
Smoker, non risk drinker | 2.7 | 0.4 | 43.9 | 52.9 | 100.0 | 2.7 | 2.1 | 47.3 | 47.9 | 100.0 |
Non smoker, risk drinker | 1.6 | 0.0 | 55.9 | 42.6 | 100.0 | 3.8 | 0.9 | 57.6 | 37.7 | 100.0 |
Non smoker, non risk drinker | 2.2 | 1.5 | 65.1 | 31.2 | 100.0 | 3.7 | 1.4 | 65.6 | 29.3 | 100.0 |
All individuals | Chi2 119.9; df 9; p < 0.001; w .24 | Chi2 65.0; df 9; p < 0.001; w .17 |
All medicament users Age 20–39 | Chi2 11.9; df 6; ns; w .09 | Chi2 5.9; df 6; ns; w .06 |
Current smokers, alcohol risk drinkers | | | | | | | | | | |
Smoker, risk drinker | 2.4 | 0.8 | 30.4 | 66.4 | 100.0 | 0.0 | 0.0 | 50.0 | 50.0 | 100.0 |
Smoker, non risk drinker | 2.1 | 0.5 | 35.2 | 62.2 | 100.0 | 1.1 | 1.8 | 38.6 | 58.5 | 100.0 |
Non smoker, risk drinker | 1.2 | 0.0 | 29.4 | 69.4 | 100.0 | 0.0 | 2.9 | 38.2 | 58.8 | 100.0 |
Non smoker, non risk drinker | 0.5 | 0.5 | 29.6 | 69.4 | 100.0 | 1.1 | 0.5 | 40.4 | 58.0 | 100.0 |
All individuals | Chi2 6.1; df 9; ns; w .09 | Chi2 6.0; df 9; ns; w .09 |
All medicament users Age 40–59 | Chi2 3.4; df 6; ns; w .12 | Chi2 5.3; df 6; ns; w .12 |
Current smokers, alcohol risk drinkers | | | | | | | | | | |
Smoker, risk drinker | 1.7 | 0.0 | 37.6 | 60.7 | 100.0 | 8.3 | 0.0 | 63.9 | 27.8 | 100.0 |
Smoker, non risk drinker | 1.3 | 0.6 | 37.9 | 60.1 | 100.0 | 4.2 | 3.2 | 51.8 | 40.7 | 100.0 |
Non smoker, risk drinker | 0.0 | 0.0 | 54.2 | 45.8 | 100.0 | 2.0 | 0.0 | 65.3 | 32.6 | 100.0 |
Non smoker, non risk drinker | 1.0 | 2.5 | 53.3 | 43.2 | 100.0 | 2.2 | 1.5 | 66.0 | 30.3 | 100.0 |
All individuals | Chi2 32.0; df 9; p < 0.001; w .20 | Chi2 18.8; df 9; p < 0.05; w .15 |
All medicament users Age 60–79 | Chi2 13.4; df 6; p < 0.05; w .16 | Chi2 11.6; df 6; ns; w .15 |
Current smokers, alcohol risk drinkers | | | | | | | | | | |
Smoker, risk drinker | 0.0 | 0.0 | 74.1 | 25.9 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 100.0 |
Smoker, non risk drinker | 6.2 | 0.0 | 70.8 | 22.9 | 100.0 | 5.0 | 0.0 | 73.3 | 21.7 | 100.0 |
Non smoker, risk drinker | 4.3 | 0.0 | 82.8 | 12.9 | 100.0 | 13.0 | 0.0 | 69.6 | 17.4 | 100.0 |
Non smoker, non risk drinker | 3.4 | 1.3 | 85.1 | 10.2 | 100.0 | 6.7 | 1.9 | 81.4 | 10.0 | 100.0 |
All individuals | Chi2 22.8; df 9; p < 0.01; w .17 | Chi2 11.6; df 9; ns; w .13 |
All medicament users | Chi2 8.4; df 6; ns; w .10 | Chi2 4.3; df 6; ns; w .08 |
Table 3
Associations between smoking and alcohol risk drinking with medicament use; multinomial logistic regression analysis
Smokers, risk drinkers | | |
Non smoker, non risk drinker | Ref | Ref |
Smoker, non risk drinker | 1.0 (0.8–1.4) | 0.7 (0.6–0.8) |
Non smoker, risk drinker | 0.7 (0.4–1.0) | 0.9 (0.7–1.2) |
Smoker, risk drinker | 1.0 (0.4–2.1) | 0.7 (0.6–0.8) |
Gender | | |
Men | Ref | Ref |
Women | 2.4 (1.6–3.5) | 1.5 (1.3–1.8) |
Age (years) | | |
20–39 | Ref | Ref |
40–59 | 2.7 (2.1–3.6) | 2.3 (2.0–2.5) |
60–79 | 18.3 (13.7–24.6) | 11.1 (9.3–13.2) |
In a further step, SO use was analyzed with respect to medical treatment. All SO users except one had contact to a physician during the past 12 months prior to the interview. Among individuals who had consulted a neurologist or a psychiatrist during the past 12 months, 10.6% took SHA and 3.4% took opioids (Table
4). Among those without a psychiatric screening diagnosis, less than 2% used SO compared to 8.2% among those with 3 or more psychiatric screening diagnoses.
Table 4
Sedatives, hypnotics or anxiolytics (SHA), opioids, and other medicaments users
Outpatient care 12 months | | | | | |
Neurologist or psychiatrist | 10.6 | 3.4 | 72.2 | 13.8 | 100.0 |
GP, not neurologist or psychiatrist | 2.2 | 1.0 | 61.9 | 34.9 | 100.0 |
Other, not neurologist or psychiatrist | 1.6 | 1.1 | 59.3 | 38.0 | 100.0 |
No | 0.0 | 0.2 | 18.8 | 81.0 | 100.0 |
All individuals | Chi2 557.9; df 6; p < 0.001; w .37 |
All medicament users | Chi2 66.7; df 6; p < 0.001; w .17 |
Inpatient care (days) | | | | | |
No | 2.5 | 1.0 | 56.2 | 40.2 | 100.0 |
1–9 | 2.4 | 2.4 | 64.5 | 30.7 | 100.0 |
10 or more | 5.9 | 2.2 | 73.7 | 18.2 | 100.0 |
All individuals | Chi2 89.8; df 6; p < 0.001; w .14 |
All medicament users | Chi2 8.7; df 6; ns: w .06 |
Psychiatric disorders (number) | | | | | |
0 | 1.3 | 0.4 | 53.4 | 44.9 | 100.0 |
1 – 2 | 2.3 | 1.1 | 58.8 | 37.8 | 100.0 |
3 or more | 5.8 | 2.4 | 64.3 | 27.6 | 100.0 |
All individuals | Chi2 133.6; df 6; p < 0.001; w .18 |
All medicament users | Chi2 47.7; df 4; p < 0.001; w .14 |
Women did not have higher odds for SHA or opioid use than men after adjustment for age, school education, income, outpatient care by a neurologist or psychiatrist, inpatient treatment, and psychiatric screening diagnoses, and using all subjects with use of other medicaments as the comparison group (Table
5). Since women and men did not differ according to the proportions of SHA or opioid users when the effect size measure Cohen's w was taken as the criterion we performed the multivariable data analysis without stratification for gender. Among individuals aged 60 to 79 higher odds were found for SHA intake when users of medicaments other than SO are the comparison group. School education of less than ten years was associated with higher odds for SHA intake than more than ten years school education when users of medicaments other than SHA or opioids were the comparison group. Individuals with 3 or more psychiatric diagnoses had higher odds for SHA or opioid intake compared to users of other medicaments.
Table 5
Associations between gender, age and sedative, hypnotic or anxiolytic and opioid medication use; multinomial logistic regression analysis
Gender | | | | | |
Men | Ref | Ref | Ref | Ref | Ref |
Women | 2.0 (1.3–2.9) | 2.0 (1.1–3.8) | 1.5 (1.3–1.7) | 1.3 (0.9–1.9) | 1.4 (0.8–2.4) |
Age (years) | | | | | |
20–39 | Ref | Ref | Ref | Ref | Ref |
40–59 | 2.5 (1.7–3.8) | 2.6 (1.3–4.9) | 2.3 (2.1–2.5) | 1.1 (0.7–1.6) | 1.1 (0.6–2.1) |
60–79 | 25.7 (19.3–34.1) | 11.5 (5.0–26.5) | 10.9 (9.0–13.3) | 2.3 (1.8–3.1) | 1.0 (0.4–2.5) |
Education (years) | | | | | |
> 10 | Ref | Ref | Ref | Ref | Ref |
10 | 1.6 (0.99–2.6) | 2.9 (1.2–7.4) | 1.0 (0.8–1.2) | 1.7 (1.0–2.7) | 3.0 (1.2–7.8) |
< 10 | 2.3 (1.5–3.6) | 2.5 (0.9–6.7) | 1.3 (1.0–1.5) | 1.9 (1.2–3.0) | 2.0 (0.7–5.8) |
Income c (Euro) | | | | | |
> 1440 | Ref | Ref | Ref | Ref | Ref |
950 – < 1440 | 0.7 (0.5–0.98) | 0.8 (0.4–1.5) | 0.8 (0.6–0.99) | 0.9 (0.5–1.5) | 1.0 (0.5–2.4) |
< 950 | 1.0 (0.7–1.4) | 1.1 (0.7–1.8) | 0.8 (0.6–0.9) | 1.4 (0.96–1.9) | 1.5 (0.8–2.7) |
Outpatient care | | | | | |
not neurologist or psychiatrist | Ref | Ref | Ref | Ref | Ref |
neurologist or psychiatrist | 8.6 (4.5–16.4) | 3.8 (2.5–5.7) | 2.6 (2.0–3.3) | 3.4 (2.0–5.7) | 1.5 (0.97–2.3) |
Inpatient care (days) | | | | | |
no | Ref | Ref | Ref | Ref | Ref |
1–9 | 1.0 (0.5–2.1) | 3.3 (1.4–7.6) | 1.5 (1.2–1.9) | 0.7 (0.3–1.4) | 2.2 (1.0–5.0) |
10 or more | 2.4 (1.4–4.2) | 3.0 (1.5–5.9) | 2.0 (1.5–2.7) | 1.2 (0.7–1.9) | 1.5 (0.9–2.7) |
Psychiatric disorders, number | | | | | |
0 | Ref | Ref | Ref | Ref | Ref |
1 – 2 | 2.3 (1.2–4.6) | 2.4 (1.2–4.5) | 1.4 (1.3–1.5) | 1.7 (0.9–3.2) | 1.7 (0.9–3.2) |
3 or more | 6.7 (3.5–13.0) | 6.5 (4.2–9.9) | 2.0 (1.7–2.3) | 3.4 (1.8–6.5) | 3.2 (2.1–5.0) |
Substance use pattern
Among men, 50.9% had one or more substance use risk behaviors, among women 35.3% (Table
6). Effect sizes were largest for the differences by age groups. Among men aged 60 to 79, there were 15.2% current cigarette smokers without SO use and 4.7% SO users in contrast to 50.8% current smokers without SO use and 2.0% SO users at age 20 to 39. Women aged 60 to 79 included 8.8% current cigarette smokers without SO use and 8.4% SO users in contrast to 41.5% current cigarette smokers without SO use and 2.1% SO users among women at age 20 to 39. According to the number of substance use behaviors among tobacco smoking, alcohol risk drinking, and SO use, 34.1% of the sample revealed one, and 8.9% two or three of the substance use risks. Among all men who practiced at least one of the 3 substance use behaviors, 6.0% used SO, and among women 14.1% used SO (Chi
2 223.1; df 3; p < 0.001; w .34). The data revealed that the proportions of individuals with at least one substance use risk behavior were lower in older than in younger age groups: among the young adults, 56.7%, among the middle adult age individuals, 46.8%, and in the older age group 26.3% disclosed at least one substance use behavior (Chi
2 281.0; df 2; p < 0.001; w .25).
Table 6
Substance use patterns
Total | 20.3 | 12.5 | 15.1 | 3.0 | 49.1 | 100.0 | 22.9 | 2.8 | 4.6 | 5.0 | 64.7 | 100.0 |
| Chi2 330.5; df 4; p < 0.001; w .27 |
Age (years) | | | | | | | | | | | | |
20–39 | 30.9 | 19.9 | 13.8 | 2.0 | 33.5 | 100.0 | 37.8 | 3.7 | 4.6 | 2.1 | 51.8 | 100.0 |
40–59 | 20.6 | 15.8 | 19.8 | 2.2 | 41.7 | 100.0 | 21.6 | 4.1 | 5.9 | 4.7 | 63.7 | 100.0 |
60–79 | 11.7 | 3.5 | 11.6 | 4.7 | 68.4 | 100.0 | 8.6 | 0.2 | 3.0 | 8.4 | 79.8 | 100.0 |
| Chi2 281.1; df 8; p < 0.001; w .36 | Chi2 249.2; df 8; p < 0.001; w .32 |
Education (years) | | | | | | | | | | | | |
< 10 | 19.2 | 9.6 | 12.4 | 3.7 | 55.0 | 100.0 | 15.7 | 1.1 | 3.2 | 7.6 | 72.5 | 100.0 |
10 | 23.4 | 16.4 | 16.9 | 2.9 | 40.4 | 100.0 | 30.7 | 3.3 | 4.2 | 3.6 | 58.2 | 100.0 |
> 10 | 14.8 | 9.5 | 17.4 | 1.6 | 56.8 | 100.0 | 15.9 | 4.9 | 10.1 | 2.3 | 66.9 | 100.0 |
| Chi2 64.2; df 8; p < 0.001; w .17 | Chi2 123.0; df 8; p < 0.001; w .24 |
Incomeb (Euro) | | | | | | | | | | | | |
< 950 | 25.3 | 17.3 | 13.8 | 3.3 | 40.3 | 100.0 | 30.1 | 4.8 | 2.8 | 4.5 | 57.8 | 100.0 |
950 – < 1440 | 20.1 | 11.0 | 16.0 | 2.6 | 50.3 | 100.0 | 22.2 | 1.4 | 5.4 | 4.4 | 66.6 | 100.0 |
> 1440 | 15.1 | 9.5 | 16.4 | 2.8 | 56.3 | 100.0 | 15.1 | 2.0 | 6.2 | 5.8 | 70.9 | 100.0 |
| Chi2 56.9; df 8; p < 0.001; w .17 | Chi2 73.9; df 8; p < 0.001; w .19 |
Outpatient care | | | | | | | | | | | | |
Neurologist or psychiatrist | 14.0 | 6.0 | 13.5 | 14.0 | 52.4 | 100.0 | 15.3 | 2.4 | 3.5 | 13.9 | 64.8 | 100.0 |
GP, not neurologist or psychiatrist | 20.4 | 11.8 | 15.1 | 2.0 | 50.7 | 100.0 | 22.9 | 2.7 | 4.3 | 4.4 | 65.7 | 100.0 |
Other, not neurologist or psychiatrist | 15.9 | 10.7 | 17.8 | 3.7 | 51.8 | 100.0 | 21.5 | 2.4 | 7.2 | 1.7 | 67.2 | 100.0 |
No | 28.1 | 21.1 | 13.4 | 0.3 | 37.1 | 100.0 | 41.3 | 4.9 | 4.9 | 0.0 | 49.0 | 100.0 |
| Chi2 108.0; df 12; p < 0.001; w .26 | Chi2 103.0; df 12; p < 0.001; w .22 |
Inpatient care (days) | | | | | | | | | | | | |
no | 20.9 | 12.8 | 15.5 | 2.6 | 48.3 | 100.0 | 23.4 | 2.8 | 4.9 | 4.5 | 64.4 | 100.0 |
1–9 | 20.4 | 13.4 | 14.1 | 2.8 | 49.3 | 100.0 | 23.0 | 4.0 | 3.4 | 6.8 | 62.8 | 100.0 |
10 or more | 15.5 | 9.3 | 11.9 | 7.2 | 56.2 | 100.0 | 17.7 | 1.2 | 2.4 | 9.2 | 69.5 | 100.0 |
| Chi2 18.1; df 8; p < 0.05; w .10 | Chi2 14.9; df 8; ns; w .08 |
Psychiatric disorders (number) | | | | | | | | | | | | |
0 | 19.9 | 12.9 | 16.2 | 0.9 | 50.1 | 100.0 | 21.9 | 2.2 | 4.1 | 2.8 | 69.0 | 100.0 |
1 – 2 | 21.7 | 11.7 | 14.4 | 2.8 | 49.4 | 100.0 | 26.4 | 1.5 | 4.6 | 4.0 | 63.6 | 100.0 |
3 or more | 18.7 | 13.1 | 13.8 | 7.9 | 46.5 | 100.0 | 20.0 | 4.7 | 5.1 | 8.3 | 61.9 | 100.0 |
| Chi2 46.8; df 8; p < 0.001; w .16 | Chi2 47.4; df 8; p < 0.001; w .15 |
Women revealed lower odds than men for substance use patterns without SO use, not however for SO use after adjustment for age, school education, income, outpatient care rendered by a neurologist or psychiatrist, inpatient treatment, and psychiatric screening diagnoses (Table
7). Therefore, we conducted further analysis without stratification for gender. Those aged 60 to 79 had an OR of 1.9 (CI 1.1–3.4) for SO use in contrast to lower odds for current smokers or alcohol risk drinkers compared to individuals who were neither current cigarette smokers nor alcohol risk drinkers nor SO users. Opposite findings for all SO users on the one hand side and smokers without SO use on the other hand side were also revealed for outpatient care provided by a neurologist or psychiatrist. Those with the lowest education had higher odds for smoking and for SO use compared to those who did not use any of the three substances. Substance use in total, current cigarette smoking, alcohol risk drinking or SO use, was lower among individuals at age of 40 or above compared to individuals younger than 40 years.
Table 7
Associations with substance use patterns; multinomial logistic regression analysis
Gender | | | | | |
Men | Ref | Ref | Ref | Ref | Ref |
Women | 0.7 (0.6–0.8) | 0.13 (0.10–0.18) | 0.2 (0.16–0.26) | 1.1 (0.8–1.5) | 0.4 (0.39–0.5) |
Age (years) | | | | | |
20–39 | Ref | Ref | Ref | Ref | Ref |
40–59 | 0.5 (0.4–0.6) | 0.7 (0.56–0.99) | 1.1 (0.9–1.5) | 1.2 (0.7–2.1) | 0.6 (0.5–0.8) |
60–79 | 0.14 (0.10–0.18) | 0.09 (0.06–0.15) | 0.5 (0.3–0.6) | 1.9 (1.1–3.4) | 0.2 (0.17–0.3) |
Education (years) | | | | | |
> 10 | Ref | Ref | Ref | Ref | Ref |
10 | 1.9 (1.5–2.5) | 1.5 (1.0–2.1) | 0.9 (0.7–1.2) | 2.2 (1.2–4.2) | 1.5 (1.3–1.7) |
< 10 | 2.3 (1.7–3.1) | 1.5 (1.0–2.3) | 0.8 (0.5–1.1) | 2.3 (1.2–4.4) | 1.5 (1.3–1.8) |
Incomed (Euro) | | | | | |
> 1440 | Ref | Ref | Ref | Ref | Ref |
950 – < 1440 | 1.3 (1.0–1.6) | 1.1 (0.8–1.5) | 1.1 (0.8–1.4) | 0.9 (0.6–1.4) | 1.1 (0.96–1.3) |
< 950 | 1.4 (1.2–1.8) | 1.6 (1.2–2.2) | 0.9 (0.6–1.1) | 1.4 (0.9–2.1) | 1.3 (1.1–1.4) |
Outpatient care | | | | | |
not neurologist or psychiatrist | Ref | Ref | Ref | Ref | Ref |
neurologist or psychiatrist | 0.7 (0.5–0.996) | 0.5 (0.3–0.9) | 0.9 (0.6–1.3) | 2.9 (2.0–4.2) | 1.0 (0.9–1.2) |
Inpatient care (days) | | | | | |
No | Ref | Ref | Ref | Ref | Ref |
1–9 | 1.0 (0.7–1.4) | 1.3 (0.8–2.1) | 1.0 (0.6–1.5) | 1.2 (0.6–2.2) | 1.0 (0.8–1.3) |
10 or more | 0.9 (0.6–1.3) | 1.0 (0.6–1.7) | 0.8 (0.5–1.3) | 1.4 (0.9–2.2) | 1.0 (0.8–1.2) |
Psychiatric disorders (number) | | | | | |
0 | Ref | Ref | Ref | Ref | Ref |
1 – 2 | 1.3 (1.1–1.5) | 0.9 (0.6–1.2) | 0.9 (0.7–1.2) | 2.0 (1.2–3.3) | 1.1 (0.8–1.7) |
3 or more | 1.1 (0.9–1.4) | 1.4 (1.0–2.0) | 1.0 (0.8–1.4) | 4.2 (2.5–6.8) | 1.3 (0.99–1.7) |
Discussion
There are four main findings. First, when taking SO use, tobacco smoking and alcohol risk drinking together, the substance-related health risk of the population is high throughout all age groups and in both genders. Second, the data suggest that there is no higher odds ratio for SO use among smokers and among alcohol risk drinkers compared to those who neither smoke nor drink alcohol in a risky way. Third, the data confirm findings from previous studies that show particularly high proportions of SO users at older age groups. Fourth, the proportion of SO users is lower than the proportion of current cigarette smokers or alcohol risk drinkers when all age groups are considered. The hypothesis is supported that in older adult age more SO use is present than in younger age groups, among men and among women. However, subjects with intake of any of the 3 substances were unequally distributed across younger, middle, and older adult age.
We did not find an association between current smoking or alcohol risk drinking with SHA or opioid use, neither in the total sample nor in single age groups. This finding indicates that there is no such relation with psychotropic medicine use in this sample as is known from smoking and risk drinking and the respective hypothesis is rejected. One explanation might be that SHA and opioid use were by far more rare than smoking in this sample. Easy availability of cigarettes and alcohol might add to that. On the other hand, particularly for elderly women a preference for SHA might be expected. However, there was no strong evidence for this.
The proportion of individuals who showed one or more risk behaviors among tobacco smoking, alcohol risk drinking and SO use is tremendously high among those younger than 40 with 66.5% among men and 48.2% among women. Substance use in older age is a prevalent health risk behavior, although the proportion of substance users at age 60 to 79 is less than half the respective proportion at age 20 to 39, both in women and men. We do not know the specific reasons for this difference. Selective mortality, smoking cessation and changes in alcohol consumption may have caused this finding.
The proportions of SHA or opioid users seem to be lower than those found in other research [
3]. This might be due to differences in data collection methods or characteristics of the samples, such as age ranges. Furthermore, our data do not confirm that the group of benzodiazepine users contains more current smokers or alcohol risk drinkers than non-users [
9]. The data confirm findings of a higher proportion of SO users in older than in young adult age [
3,
8] and they compare with data that revealed high proportions of benzodiazepine consumers among the elderly [
2]. However, no differences by age were present when SO users are compared to individuals with other medicament intake except for SHA at age 60 to 79. The adjusted OR for SO users compared to individuals without smoking, without alcohol risk drinking and without SO use is 1.9 with a lower confidence bound close to 1 (CI 1.1 to 3.4). The bivariable statistics revealed higher proportions of individuals with SHA or opioid use in higher age groups among all medicament users only for women. Even in older adult age, smoking and alcohol risk drinking are the main substance use risk behaviors among men whereas among women the proportion with SO use is close to the proportion of current smokers and alcohol risk drinkers.
Less educated individuals and those with the lowest income had higher odds for current cigarette smoking and for SO use after adjustment compared to those with the highest school education and the highest income respectively. The data suggest that proportions of SO users are higher among those with a larger number of psychiatric disorders, not however the proportions of individuals with other SP.
Substance use patterns are different with respect to gender, age, education, and utilization of outpatient medical care. Women have substantially less alcohol risk drinking whereas they do not differ from men according to SO use. While there are less current smokers, with or without co-occurrent alcohol risk drinking, there are more SO users at age 60 – 79 compared to those younger than 40. Several explanations might hold for this finding. On the one hand side concerns about one's health status might be responsible for quitting smoking and alcohol risk drinking among the older individuals, whereas on the other hand problems of quality of life at older age, such as how to cope with sleeplessness, might stimulate SO use. However, there are differences by education and income indicating that individuals with less education presented higher proportions of smokers, alcohol risk drinkers and SO users. One reason might be that substance use and morbidity is higher in subpopulations with low compared to those with high socioeconomic status. One subgroup among SO users might be those who are in neurological or psychiatric outpatient treatment.
Our data do not allow conclusions according to non-medical or medical SO use. The data revealed more SO use among individuals who had consulted a neurologist or a psychiatrist during the past 12 months prior to the health examination than among individuals without service use. However, even among individuals in medical treatment, there may have been non-medical use if they took the substances in larger amounts than prescribed. Moreover, we could not identify the reasons for taking the substance including use according to prescription only, abuse or dependence on prescription drugs in conjunction with illicit drug use or prescription drug use when illicit drugs of choice were not available [
21].
One objection against our measurement of substance use patterns is that current smoking, alcohol risk drinking and SO use may be comparable only in part. Differences in availability of the substances must be considered. Neither exists a standardization of the amount of toxic effects which the individuals are exposed to across tobacco smoke, alcohol and SO nor is there evidence of a standard score according to their risk potential. Diseases caused by tobacco smoking and alcohol risk drinking are largely known, and there also exists evidence about dose-response-relations. We know less from SO use. Our data did not include how long and in which dose SO have been taken. Guidelines for risks are available for alcohol consumption [e.g. [
14]] whereas for SO use it seems to be largely unknown which lower limit exists for attributable morbidity and mortality depending on the single drug. There is some knowledge about the risk increase by the co-occurrence of tobacco smoking and alcohol risk drinking but we hardly know anything about how SO use interacts with consumption of the other two substances. Furthermore, tolerance and total lifetime doses of exposure until manifestation of disease that is attributable to the substance use may differ between men and women.
Further limitations of our study are: (1) The sample was drawn in one region of Germany only. It is not representative for the whole country due to its low population density and different social structure. However, we assume that this does not change the associations between smoking alcohol risk drinking and potential confounders with SO use as analyzed in this study. (2) No data about current illicit drug use were available. (3) The sample includes only individuals who were younger than 80 years when the sample was drawn. Above that age there seem to be less current smokers and alcohol risk drinkers whereas SO use may be more prevalent than in the age range of our sample [cf. [
3,
11,
22]]. (4) In several cases the cell numbers were small. Thus, more significant differences may be expected in a larger sample for SHA and for opioid use for age, education and income, outpatient and inpatient care, and number of psychiatric disorders (5) This is a cross-sectional study. Thus, age cohort effects cannot be excluded. (6) For each of the three substance use behaviors different reporting bias may have been active, which is expected high for alcohol drinking amounts and low for smoking and SO use. The reason is that there were low health policy efforts to combat smoking and SO use at the time of the data collection or before that time. A strength of the present study is the data gathering of medicaments. This might have added to the reduction of reporting bias. But we may have missed some drugs. Recall bias may have been active among those who did not bring their medicament packages. Even among those who brought their drugs, some may have missed single packages. Individuals who had prescriptions actually may have not have taken the medication. (7) We do not know the duration of medicament intake. (8) Beyond drugs, the data have been gathered on grounds of self-statements and were not verified biochemically. However, other evidence suggests that in population-based studies, there is no considerable risk of denial which significantly influences the results [
23].