Pain and alcohol: a comparison of two cohorts of 60 year old women and men: findings from the Good Aging in Skåne study

1 Introduction

A striking conclusion of the World Health Organisation (WHO) is that the inhabitants of the European Region consume the most alcohol per capita in the world [1]. It is now widely known that alcohol interacts with medication and can lead to serious health problems. Merrick and colleagues found that nine percent of elderly individuals reported heavy drinking (i.e. more than 168 g alcohol/week for men and more than 108 g alcohol/week for women [2]) despite having a higher educational level and better health status than those who did not report excessive alcohol consumption [3].

New findings suggest that the recommended risk thresholds for alcohol consumption should be lowered as 100-g or more of alcohol per week may linearly lower life expectancy [4]. In most cases alcohol-related problems refer to those who are dependent on alcohol. Alcohol may cause problems among older individuals, regardless of how many grams they consume, because of their changed physiology and the interaction with painkillers [5].

Older adults change their drinking patterns in diverse ways after their health deteriorates [6]. As health, social and economic aspects affect older persons’ drinking habits it is important to consider their whole situation when taking measures to reduce alcohol consumption [6], especially among those who use painkillers. All painkillers interact with alcohol, for example the level of glutathione decreases after ingestion of therapeutic doses of acetaminophen, which may increase the risk of hepatocellular necrosis [7].

It has previously been reported that more than 50% of the elderly population (60 years and older) have pain [8]. The use of prescription painkillers has been reported by almost 60% of women and nearly 50% of men aged 60 years and older [8]. This possible association between pain and alcohol is of increasing empirical interest because alcohol and analgesics often interact with each other and alcohol also has an analgesic effect. Pain at an older age entails an elevated risk of alcohol abuse [9]. Therefore, clinicians need to be observant of alcohol consumption among their pain patients – especially men [10].

A study of accidental mortality associated with acetaminophen and codeine use found that nearly 80% of these deaths involved another medication or alcohol [11]. It is also known that pain is common in individuals seeking help for alcohol addiction [12]. Gender differences were revealed in the study of drinking habits where daily alcohol intake in women was associated with better mental health, while men’s daily drinking was associated with better physical health [13]. Patients treated with medical cannabis often use alcohol in a manner that poses a high risk [14]. Patients with mild to moderate pain are at higher risk of using alcohol in an unhealthy way compared to patients reporting severe pain [14]. Laboratory findings suggest that acute pain is related to intentions to drink but no relationship with the demand for alcohol was found [15]. Consequently, improving pain-coping skills may result in a decrease in coping-motivated drinking [15].

The aims of this study were to describe differences in pain, alcohol consumption and use of painkillers among two 60 year old birth cohorts stratified by gender.

2 Methods

3 Results

Men in the two cohorts had a mean age of 60.3 years and a majority were married or cohabiting (79.4% and 75.8%). Men born between 1953 and 1954 more often lived in rural areas (p<0.004), see Table 1.

Men in the 1941–43 cohort had a lower educational level (elementary school) p<0.007. Men in the 1952–54 cohort reported pain more often, 281 (59.0%), than men born between 1941 and 1943, 173 (51.6%), p<0.034. No difference in use of painkillers was found between men in the two cohorts, 69 (39.9%) vs. 88 (31.3%), p<0.062. Teetotallers or those who seldom consumed alcohol were more common among men in the 1952–54 cohort, 128 (29.0%) vs. men born between 1941 and 1943, 92 (27.5%), p<0.029. No difference in alcohol use was detected between the two male cohorts; 166 g/month (q1 36–q3 348) vs. 133 g/month (q1 26–q3 282), p<0.134, even when divided into different pain groups, see Table 1.

Women in the two cohorts had a mean age of 60.3 and 60.2 (n.s.). A majority were married or cohabiting but women in the 1941–43 cohort more often reported being single/widowed/divorced, 123 (37.5%) vs. 126 (28.1%), p<0.006. Women in the 1952–54 cohort more often lived in rural areas, 27 (6.6%) than women in the 1941–43 cohort, 16 (4.9%), p<0.001, see Table 2.

Women in the 1941–43 cohort more commonly had a lower educational level (elementary school) 122 (37.3%) vs. 87 (21%), p<0.001, but no difference was found between the cohorts in terms of pain, regardless of intensity, p<0.144. Women in the 1941–43 cohort more often used painkillers, 110 (53.1%) vs. women born in the 1952–54 cohort 119 (40.1%), p<0.004. Teetotallers or those who seldom used alcohol were about the same in the two female cohorts, 138 (42.2%) vs. 158 (38.1%), p<0.662. No difference in alcohol use was detected between the two female cohorts, 81 g/month (q1 11–q3 227) vs. 76 g/month (q1 8–q3 182), p<0.225, even when divided into different pain groups, see Table 2.

The percentage of men with severe pain who used painkillers was 61.1% for the 1941–43 cohort and 73.3% for the 1952–54 cohort but with no difference between the cohorts (p<0.376). It was also common for men with severe pain to combine painkillers and alcohol, as this applied to eight men in the 1941–43 cohort (44.4%) vs. 13 men in the 1952–54 cohort (44.8%), with no difference between cohorts (p<0.980), see Table 3.

When comparing the two female cohorts, it was found that the use of painkillers is more common among women in the 1941–43 cohort, 110 (53.1%) vs. 119 (40.1%), p<0.004, who more often reported alcohol consumption than women in the 1952–54 cohort, 80 (38.8%) vs. 82 (28.9%) p<0.021. The combination of painkillers and alcohol is most frequently reported by women with moderate pain: Women in the 1941–43 cohort, 45 (48.9%) vs. women in the 1952–54 cohort, 44 (39.3%), but without a significant difference between the cohorts, p<0.202, see Table 4.

When comparing men and women with pain in the 1941–42 cohort, men with moderate pain use more alcohol than women, 157 g/month (q1 10–q3 365) compared to 44 g/month (q1 0–q3 134), p<0.001. Men with severe pain also use more alcohol, 96 g/month (q1 17–q3 324) than women, 27 g/month (q1 0–q3 118), p<0.030 and when those with pain are merged, the men use more alcohol 175 g/month (q1–31–q3 356) than the women 68 g/month (q1 1–q3 207), p<0.001, see Table 5.

When comparing men and women in the 1952–54 cohort, men with no pain use more alcohol than women 137 g/month (q1 34–q3 285) compared to 76 g/month (q1 9–q3 188), p<0.005, men with little pain, 147 g/month (q1 45–q3 285), women with little pain 96 g/month (q1 23–q3 2017), p<0.025, men with moderate pain, 107 g/month (q1 11–q3 300), women with moderate pain 70 g/month (q1 0–q3 181), p<0.024, and when all individuals with pain are merged in one group, men 132 g/month (q1 22–q3 270), women 76 g/month (q1 8–q3 182), p<0.001. The only exception is the group reporting severe pain where no gender difference could be detected, p<0.175, which differs from men and women in the 1941–43 cohort, where men reporting severe pain use more alcohol, 207 g/month (q1 94–q3 407) than women 91 g/month (q1 17–q3 176), p<0.045, see Table 5.

When all reported pain was merged into one group a significant gender difference was found where women used more painkillers. In the 1941–43 cohorts 69 men reported using painkillers (39.9%) compared to 110 women (53.1%), p<0.010, while in the 1952–54 cohorts 88 men reported using painkillers (18.5%) compared to 119 women (26.6%), p<0.003. A gender difference was found when comparing different pain groups and use of painkillers and alcohol. In the 1941–43 cohort women used the combination more often (80, 38.6%) than men (48, 27.7%), p<0.025. However, when all pain was merged into one group and alcohol reported in g, a significant gender difference was found between both cohorts. In the 1941–43 cohort men reported 190 g/month (q1 94–q3 345) vs. women 113 g/month (q1 48–q3 228), p<0.008, while in the 1952–53 cohort men reported consuming 210 g/month (q1 89–q3 365) vs. women 113 (q1 47–q3 2012, p<0.005), see Table 5.

It no longer applies that those experiencing pain are more restrained with alcohol than individuals without pain except for women with pain in the 1941–43 cohort who used significantly less alcohol than women without pain; men with pain in the 1941–43 cohort 175 g/month (q1 31–q3 356) vs. those without pain 160 g/month (q1 38–q3 330), p<0.798; men with pain in the 1952–54 cohort 132 g/month (q1 22–q3 270) vs. without pain 137 g/month (q1 34–q3 285), p<0.610; Women with pain in the 1941–43 cohort 68 g/month (q1 1–q3 207) vs. those without pain 114 g/month (q1 29–q3 286), p<0.008; and women with pain in the 1952–54 cohort 76 g/month (q1 3–q3 182) vs. those without pain 76 g/month (q1 9–q3 188), p<0.630.

Among men in the 1941–43 cohort, eight (44%) with severe pain reported that they used both alcohol and painkillers. This 44% used alcohol to a higher extent than other men reporting pain in this cohort and both the mean and quartiles are elevated; 207 g/month (q1 94–q3 407), but without a corresponding increase in alcohol intake in women with severe pain 91 g/month (q1 17–q3 176), p<0.045.

4 Discussion

This study compares cohorts of the same age (60 years) at a 12 year interval to investigate whether alcohol habits changed over time, especially in the section of the population that has pain and uses pain medication. No difference in drinking behavior between the two male cohorts was found, except that men in the 1953–54 cohort were more often teetotallers. The two female cohorts showed no difference in alcohol intake. There was no increase in the number of teetotallers or those who rarely used alcohol among women in the 1953–54 cohort, which is in line with previous research showing that the number of female teetotallers in Sweden decreased between 1997 and 2001 [17].

Men with severe pain seem to use less alcohol than those reporting little or moderate pain, which is in line with findings among the women and corresponds well with previous research [14]. The women in the 1941–43 cohort use painkillers and alcohol to a higher extent than those in the 1953–54 cohort. This should underline the need for a new approach where older women with pain receive the greatest attention in discussions about alcohol, especially when prescribing pain medication, which is in contrast to previous research where the main focus was on men’s alcohol habits [10]. Although there are gender differences in the amount of reported alcohol consumption in the two cohorts, both men and women with pain who use analgesics indicate a mean alcohol intake that is well below the stated risk level. However, the combination of analgesics and alcohol makes them a potential risk group. Alcohol consumption has been stated in grams, but consideration should be given to the fact that women usually weigh significantly less than men and thus may be at the same or perhaps even greater risk than men from the reported alcohol intake. The present findings show no difference in alcohol use among individuals with or without pain except for women in the 1941–43 cohort, thus contradicting previous research revealing that pain patients drank alcohol both in smaller quantities and more rarely compared to healthy controls [18]. A possible explanation for the differences is that the study by Thelin Bronner et al. was performed on a sample from a pain clinic. The present study investigates alcohol habits in two cohorts of 60 year old women and men from the general population with focus on those suffering from pain and using analgesics. The average alcohol consumption is estimated to be below the risk limit in both cohorts regardless of gender, which may be due to an underestimation of the amount of alcohol consumed as the results are not consistent with previous surveys of alcohol consumption [3]. Men with severe pain who use painkillers in the 1941–43 cohort report the highest alcohol intake, where the third quartile indicates that although some individuals have a high alcohol consumption, they are still below the risk limit.

Men consume more alcohol and have long been regarded as the major risk group for mixing alcohol and analgesics. While we must continue to ensure that men reduce their alcohol intake, the comparison between these cohorts indicates a hopeful new trend that future generations of men are becoming more cautious about alcohol as more men in 1953–54 cohort are teetotallers or rarely consume alcohol.

Some limitations must be mentioned. A previous study has shown that alcohol consumption may vary according to the measure used among individuals with depression who reported heavy episodic drinking [19]. In the present study, alcohol consumption is given in grams consumed without knowledge of whether it concerns daily intake or only alcohol consumed on special occasions. To clarify the differences in consumption within the groups the first and third quartiles have been indicated in the tables. Some of the groups are very small and therefore the results should be interpreted with caution. The results nevertheless indicate alcohol use in society and to the best of our knowledge, this is the first study in which cohorts of the same age have been compared gender wise in a 12 year follow-up with focus on pain, analgesics and alcohol use. Alcohol consumption is also self-reported, which may mean that the actual consumption is higher. This is substantiated by previous research [3] showing that 9% consume alcohol in excess of the risk limit, which differs from the present findings. Perhaps a computer-based application might improve reporting in the future [20], although the data will still be self-reported despite being collected electronically. The investigations were conducted at 12 year intervals, which in some cases means that the research staff were changed but special training was provided to minimize sources of error. The interval between the investigations also gives rise to some potential confounding variables as the cohorts differ regarding educational level, residential area and marital status. However, the strength of the study is the large number of individuals from two different cohorts randomized from a general population and that a test-retest has been conducted for the questions about alcohol, pain and use of painkillers.

Pain and alcohol use are both highly prevalent in the general population and the pain-alcohol relationship is of increasing empirical interest, particularly for clinicians prescribing painkillers, as the combination could be dangerous, especially in combination with older age where liver and kidney function are decline. Most of the disability-adjusted life years in the Nordic countries between 1990 and 2013 were due to premature mortality caused by alcohol [21]. Research has shown similar patterns in Scotland during the late nineties and early 2000s, where alcohol-related mortality in the 40–70 year age group increased markedly [22]. Drinking habits can be influenced by poor health [6]. There is still very little evidence that the measures taken in various countries have had an impact on alcohol consumption [23]. Selective alcohol prevention programmes report successful outcomes among specific groups [24], so reaching the risk groups, especially those on painkillers, with tailor-made interventions should be highly prioritized.

The issue of pain medication needs to be highlighted. It is obviously problematic that so many women and men consume alcohol in combination with analgesic drug therapy. Cannabinoids have become an increasingly common drug in pain management (both neuropathic and non-cancer pain) and are often combined with opioids. While the patients who try cannabis are mainly younger, about every sixth pain patient has tried it [25]. Combination treatments with several analgesics and alcohol are becoming more common, even though we do not know how they affect the individual in the long term. An equally big problem is that many men are not adequately treated for their pain or associated problems, which in many cases involve disturbed sleep [26]. Previous studies have shown that pain greatly reduces women’s quality of life [26]. Among men, sleep problems cause the lowest quality of life [26]. This may partly explain the high alcohol consumption of both men and women who report pain. When treating the elderly with pain it is important that not only the pain and medication is investigated and validated but also sleep and alcohol habits. A holistic approach is desirable.

It is thus high time to create a model for a lasting reduction of alcohol intake in the population. New findings show that the recommendations for risk consumption of alcohol should be reduced as it shortens life expectancy [4]. Better pain management strategies can reduce the desire to drink, but a lifestyle change strategy that can be implemented at health centres is important. A problem with the evaluation is that alcohol consumption is necessarily self-reported and therefore the amount can be inaccurate, but long-term follow-up with the registration of alcohol-related injuries and hospitalizations can be a possible way to evaluate efforts to influence lifestyle and alcohol consumption patterns. Perhaps the answers would be more reliable with computer-based applications because, according to the authors, it is possible to circumvent cultural taboos around alcohol by employing this model [20].

5 Conclusions

This study highlights the fact that many pain patients; men with severe pain born between 1941 and 1943 (44.4%); men with severe pain born between 1952 and 1954 (43.3%); women with moderate pain born between 1941 and 1943 (48.9%); women with moderate pain born between 1952 and 1954 (37.6%), use alcohol in combination with analgesics.

6 Implication

It is time to alert prescribers on the risk behavior of a large section of the population. This increased awareness also necessitates some form of action. It is important that the elderly obtain optimal pain relief, but it is equally vital to reduce alcohol consumption among older individuals who also use analgesics [4, 5]. More research is needed to better understand the implications of using both painkillers and alcohol in the long-term and to find strategies to reduce drinking.