How prevalent and severe is addiction on GABAmimetic drugs in an elderly German general hospital population? Focus on gabapentinoids, benzodiazepines, and z‐hypnotic drugs

1 INTRODUCTION

A lively debate is emerging concerning the abuse, drug-related harms, and addictive potential of the gabapentinoids (GPT) gabapentin and pregabalin which have been increasingly prescribed worldwide for both labeled and off-label indications (Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; McAnally et al., 2020; Smith et al., 2015). National representative epidemiological surveys on substance use have not incorporated investigations of GPT so far (e.g., Seitz et al., 2019).

1.1 GPT share important liabilities with other GABAmimetics: Misuse and respiratory depression

Increased risk for GPT misuse has been proposed for patients with psychiatric disorders (Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021) although that proposal remains understudied, as has the rate and risk of GPT abuse among the general population (Bonnet & Scherbaum, 2017). Abuse of GPT does appear to be much more prevalent among, and primarily confined to individuals with other substance use disorders (SUD), in particular opioid use disorder (Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; McAnally et al., 2020; Smith et al., 2015). Although GPT display weaker Central Nervous System (CNS)/respiratory depression than benzodiazepines (BDZ) and z-hypnotic drugs (ZD) (Bonnet and Scherbaum, 2017, 2018; Webster & Karan, 2020), GPT abuse has recently been shown to occur in the context of overdose deaths in these special populations (Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; Smith et al., 2015). Benzodiazepines, ZD and GPT prescriptions have been all demonstrated to be associated with overdose mortality in illicit opioid users and in those receiving Opioid Maintenance Therapy (Abrahamsson et al., 2017; Gomes et al., 2017, 2018; Torrance et al., 2020), suggesting potential synergism between opioids and GABAmimetic drugs – including GPT – in terms of respiratory depression.

1.2 Addiction biology of GABAmimetics

Benzodiazepines and ZD are typical GABAmimetic drugs, which unlike GPT exert pro-agonistic activity at certain brain and peripheral transmembrane GABA-A receptors by occupying special binding sites on those receptors and increasing anionic transport via allosteric modulation (Atkin et al., 2018; Crestani et al., 2000; Rourke & Law, 2021). While GPT do not bind GABA receptors but rather the alpha-2-delta subunit receptor of voltage-gated calcium channels, and have been widely held to exert no GABA agonism (Patel & Dickenson, 2016; Sills & Rogawski, 2020). Gabapentinoids have been shown to increase cerebral and spinal extracellular GABA concentrations in a dose-dependent manner in lower species (Czuczwar & Patsalos, 2001; Errante et al., 2002; Honmou, Kocsis, & Richerson, 1995; Honmou, Oyelese, & Kocsis, 1995; Kuzniecky et al., 2002; Petroff et al., 1996, 2000), including within the ventral striatal nucleus accumbens (Peng et al., 2008), a key area of the mesolimbic dopaminergic reward system, and also within the human visual cortex (Cai et al., 2012). This GPT-induced extracellular synaptic and extrasynaptic GABA-enhancement (which mechanism is discussed in Bonnet et al., 1999), may complement the well-known inhibitory action of GPT on presynaptic N and P/Q-type voltage-gated calcium channels. The latter mechanism is known to “throttle down” neuronal activity-dependent excitatory neurotransmission, thereby most likely attenuating central/behavioral sensitization, a CNS-plasticity state hypothesized to be crucial for chronic pain and addictive behavior development (Berridge & Robinson, 2016; Bonnet & Scherbaum, 2017; Cao et al., 2013; Kurokawa et al., 2011; McAnally et al., 2020; Volkow & Morales, 2015). This mechanism could be amplified by an increase in extracellular GABA-concentration (“ambient” GABA) which is attributed to an increase in tonic inhibition (by activating extrasynaptic GABA-A receptors) with impact on neuronal excitability/plasticity, mood regulation, sleep, memory and cognition, as well as pain (decreasing hyperalgesia) (Perez-Sanchez et al., 2017; Walker & Semyanov, 2008).

Physical dependence signs including withdrawal phenomena (seizures, delirium, insomnia, anxiety, agitation), as well as intoxication symptoms (euphoria, relaxation, somnolence, coma) resembling that of BDZ (Soyka, 2017) also suggest clinically relevant GABAmimetic activity with GPT (Bonnet & Scherbaum, 2017).

By the current neurobiological paradigm, the development of addictive behavior (Berridge & Robinson, 2016; Korpi et al., 2015; Volkow & Morales, 2015) and disruption of behavioral control (Deserno et al., 2015) should be accompanied by ventral striatal dopaminergism, which has been shown with BDZ in the rat brain (Tan et al., 2010), but refuted for GPT (Coutens et al., 2019; Vashchinkina et al., 2018). Nonetheless, drug-seeking behavior was recently described in a murine model exposed to hypertherapeutic gabapentin doses (Althobaiti et al., 2020).

1.3 Clinical comparison of GPT with other GABAmimetics in addiction medicine

Clinical studies comparing the addictive power of GPT with that of other substances of abuse remain a rarity (Bonnet et al., 2019; Coβmann et al., 2016; Rourke & Law, 2021); such investigations are required to gain better understanding of both the prevalence and potential severity of GPT-dependence and addiction as well as the role of substance-specific and other contextual factors that influence its development and maintenance. One of the few studies to date to examine this issue (Coβmann et al., 2016) collected data about abuse and dependence upon GPT and other non-opioid analgesics (NOA), as well as BDZ and ZD. While this investigation targeted an elderly hospitalized population, nearly identical prevalence rates of dependence upon many substances (most notably alcohol and nicotine) were seen in comparison to a more representative German adult general population (Pabst et al., 2013). Post-hoc analysis of the former sample revealed unexpected degrees of both physical and psychological dependence upon NOA (especially non-steroidal anti-inflammatory drugs [NSAIDs] and metamizole, both commonly believed to possess little dependence and addiction liability) (Bonnet et al., 2019).

1.4 Rationale of this study

Given this biological and phenomenological backdrop, we hereby present an initial clinical comparison of the dependence and addiction liability of GPT with BDZ and ZD in humans, also derived via post-hoc analysis of the original population.

2 MATERIALS AND METHODS

This study comprises a post-hoc analysis of a previous cross-sectional study on substance-dependence/addiction among a random sample of elderly German metropolitan general hospital patients. The original study was carried out (in accordance with the guidelines of the Declaration of Helsinki, and approved by the ethics commission of the Medical Association Westfalen-Lippe, North Rhine-Westphalia, Germany and the Medical Department of Westfälische Wilhelms-University of Munster, Germany) from spring to fall in 2013 in the Evangelisches Krankenhaus Castrop-Rauxel (Coβmann et al., 2016). Development of the original study sample along with inclusion and exclusion criteria is shown in Figure 1.

image

Sample realization of the original study (above the dashed line, Coβmann et al., 2016) with unfolding the population of the post-hoc analysis presented here. Inclusion criteria: inpatients of the Evangelisches Krankenhaus Castrop-Rauxel aged ≥65 years, able to grasp the explanation of the study and to voluntarily give their written informed consent; all departments were considered with exception of the intensive care unit. Exclusion criterion: Mini-Mental Status Test of <25 points (pointing to cognitive impairment). In this figure, the relative frequencies above the dashed line refer to the whole hospital population in question (N = 2801, retrospectively determined). The relative frequencies below the dashed line refer to all patients finally examined with SKID-I (N = 400; Wittchen et al., 1997). They express the geriatric hospital population's lifetime prevalence of the dependence on GABAmimetic drugs (“raw” condition, see methods). *after simple randomization; **according to SKID-I (Wittchen et al., 1997)

Originally, 671 inpatients ≥65 years old were visited in various departments (General and Trauma Surgery, 29% of the sample; Internal Medicine, 24%; Neurology, 18%; Geriatrics, 13%; Psychiatry, 8%; Gastroenterology, 6%; Gynecology, 2%) after their third hospital day, assuming that at that time, the acute admitting condition was on the path of improvement (Coβmann et al., 2016). A simple randomization measure was applied at admission to create the initial sample and from these 671 randomly selected patients, 224 refused participation and 46 were not included due to cognitive impairment (<25 points scored on the Mini-Mental Status Test). During the study one patient withdrew his consent for study participation. Hence, a cohort of 400 elderly inpatients (mean age = 75 ± 6.4 years old; 63% females) was included for examination as to current and past abuse of and dependence upon different substances (Figure 1, Coβmann et al., 2016).

Data were collected via SKID-I, a well-tested structured clinical interview identifying psychiatric disorders based on DSM-IV-TR (APA, 2000; Wittchen et al., 1997); the original interviews were conducted face-to-face by Dr. Johanna-Cristina Strasser (née Coβmann), a psychologist. SKID-I represents the German equivalent to SCID-I (First et al., 1996); this instrument delivers the highest standard in the area of clinical interview diagnostics in Germany (Braehler et al., 2002; Strauβ & Schumacher, 2005) supplying among other things measures of the severity, development and course of psychiatric diseases (Wittchen et al., 1997). The Addiction Section (E) of SKID-I can be used to determine lifetime and current 12-month prevalence rates of abuse and dependence upon various substances, and for the original study the database was expanded to include GPT, NSAIDs, acetaminophen, and metamizole (Coβmann et al., 2016). Past dependence (referred to as dependence in remission [Wittchen et al., 1997]) data are relevant in terms of assessment of the chronification/sustainability of a substance's dependence liability. Current and past dependence states are used to describe the lifetime prevalence of a condition.

In this post-hoc analysis, we re-evaluated the sample's SKID-I interview data (see below) on abuse and dependence upon BDZ, ZD (zapeclon, zolpidem, zopiclone) or GPT. Descriptive statistics and t-tests were performed in Excel. Groups were compared using a chi-square (χ2) test with Monte Carlo Simulation (because some groups contained <5 cases). In all tests, the significance level was set at p < 0.05.

2.1 Describing abuse, dependence and its severity

SKID-I uses the abuse and dependence criteria of DSM-IV-TR. Substance abuse was diagnosed if one (or more) of the following symptoms occurred within a 12-month period: (1) recurrent substance use resulting in a failure to fulfill major role obligations at work, school, or home; (2) recurrent substance use in situations in which it is physically hazardous; (3) recurrent substance-related legal problems; (4) continued substance use despite having persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of the substance. Substance dependence was diagnosed if three (or more) of these symptoms are occurring within a 12-month period: (1) tolerance; (2) withdrawal symptoms; (3) the substance is often taken in larger amounts or over a longer period than was intended; (4) Persistent desire or unsuccessful efforts to cut down or control substance use; (5) a great deal of time is spent in activities necessary to obtain the substance, use the substance, or recover from its effects; (6) important social, occupational, or recreational activities are given up or reduced because of substance use; (7) the substance use is continued despite knowledge of having a persistent or recurrent physical or mental problem that is likely to have been caused or exacerbated by the substance. Physical dependence was diagnosed if symptoms 1 and/or 2 were present. Psychological dependence was derived from symptoms 3 and 4.

Substance abuse or dependence was subdivided into three types: (1) “Raw” abuse or dependence condition, comprising cases involving the drug in question either in isolation or with concurrent substance abuse/dependence; (2) “Mono” abuse or dependence condition, comprising cases involving only the drug in question without other concurrent substance abuse or dependence; and (3) “De novo” abuse or dependence condition, comprising cases where the drug in question represented the index/first-ever substance of abuse or dependence. A special question evaluating the latter condition was added to the interview. The prevalence and severity of “mono”- and “de novo” substance dependence for a given drug are considered to correlate well with its (substance-specific) addictive power (Bonnet & Scherbaum, 2017; Bonnet et al., 2019). According to SKID-I (Wittchen et al., 1997) the severity of the dependence was assessed as follows: mild dependence (presence of 3 symptoms as well as minor impairment of occupational performance, social relationship skills and activities); moderate dependence (between mild and severe); severe dependence (above 3 symptoms as well as profound impairment of occupational performance, social relationship skills and activities).

3 RESULTS

None of the participants were admitted for a substance abuse or dependence condition, and there was no abuse of/dependence upon illicit drugs among the 400 participants (Table 1). SKID-I interviews revealed neither current nor past abuse of BDZ, ZD or GPT; 55 (13.75%) persons were found to be currently or previously dependent on BDZ, ZD and/or GPT (Figure 1). BDZ dependence was most prevalent in this sample, followed by ZD, and finally GPT (Tables 1 and 2, Figure 1). The vast majority of dependence conditions for any substance were mild-to-moderate (Figure 2, Tables 1 and 2).

TABLE 1. Substance abuse and dependence (“raw” condition, see methods): Prevalence rates and co-use; total population (N = 400)a image TABLE 2. Characteristics of hospitalized elderlies and their substance dependences image image

Profiles of the dependences on GABAmimetic drugs, describing their prevalence rates (in brackets) and severities. In contrast to benzodiazepines and z-hypnotic drug, mono-dependence on gabapentinoids (GPTs) was not found (red box) and de novo dependence of GPTs were mixed with other de novo substance dependences (see Section 3.3.3, Table 2). 12-Month-D, 12-month prevalence – (raw) dependence; 12-Month-De Novo, 12-month prevalence – de novo-dependence; 12-Month-MD, 12-month prevalence – mono-dependence; LTP-D, lifetime prevalence – (raw) dependence; LTP-D-De Novo, lifetime prevalence – de novo-dependence; LTP-MD, lifetime prevalence – mono-dependence

As per the Figure 1 legend, the prevalence rates described below refer to the 400 patients included into the original investigation (Coβmann et al., 2016). Table 2 shows characteristics of the present post hoc sample, including demographic and clinical data.

3.1 BDZ dependence 3.1.1 Raw dependence

Twenty eight of 400 patients were currently or previously dependent on BDZ resulting in a lifetime prevalence of 7% (Figure 2). The following BDZ were involved: diazepam (n = 7), lorazepam (n = 6), oxazepam (n = 4), bromazepam (n = 3), temazepam (n = 3), flurazepam (n = 2), clonazepam (n = 1), lormetazepam (n = 1), dipotassium clorazepate (n = 1). One person experienced dependence upon two BDZ.

Nineteen patients had current or previous co-use of other substances. Current co-use comprised alcohol (n = 4), nicotine (n = 4), opioid analgesics (n = 8) and NOAs (n = 4); previous co-use comprised alcohol (n = 5), nicotine (n = 5), opioid analgesics (n = 2), NOAs (n = 2) (Table 1). All patients were physically and psychologically dependent on these drugs, except for four who reported no signs of psychological dependence on BDZ.

Eleven (39.3%) of the 28 BDZ dependent patients had been abstinent from BDZ at the time of the interview; 17 (60.7%) were currently dependent on BDZ, resulting in a 12-month prevalence of 4.25% (Table 2, Figure 2). Among these, concurrent opioid analgesic dependence (n = 8) and NOA-dependence (n = 4) were reported (Table 1), with prior co-occurring dependence upon alcohol and nicotine also occurring with some frequency (n = 4, each). All patients were physically and psychologically dependent upon these substances as well as upon BDZ. Only two of these patients exhibited severe dependence, with the remainder exhibiting mild or moderate dependence (Table 2, Figure 2).

3.1.2 Mono-dependence

Among all 28 BDZ-dependent patients, 6 (21.4% of this group) and 3 (10.7% of this group) exhibited previous or current mono-dependence on BDZ, respectively (lifetime prevalence: 2.25%, 12-month prevalence: 0.75%). Only two of these patients exhibited severe dependence, with the remainder exhibiting mild or moderate dependence (Table 2, Figure 2). Of note, BDZ mono-dependence emerged significantly earlier than ZD mono-dependence (p < 0.001), on average during the sixth decade of life (Table 2).

3.1.3 De novo-dependence

Eleven persons (39.3% of all BDZ-dependents, mean age 73.7 ± 5.4 years old; 10 of them females, Table 2) reported BDZ as their index substance of abuse or dependence (lifetime prevalence: 2.75%). Among these, the dependence was rated mild to moderate, both physically and psychologically. Four patients (36.4% of this group) were in stable remission; their BDZ-dependence (diazepam n = 2, lorazepam and oxazepam, each n = 1) had lasted 1–4 years (1.75 ± 1.5 years). Seven patients were currently dependent (12-month prevalence: 1.75%) (Table 2, Figure 2). Benzodiazepines de novo-dependence emerged significantly (p < 0.001) earlier than ZD dependence, on average during the fifth decade of life (Table 2).

3.2 ZD dependence 3.2.1 Raw dependence

Seventeen patients were currently and previously dependent upon ZD (lifetime prevalence 4.3%) (Table 2, Figure 2), with 14 dependent upon zopiclone (82.4% of this group) and, the remainder on zolpidem. No person reported remission of ZD-dependence (12-month prevalence: 4.3%; Table 2, Figure 2), with all patients physically and psychologically dependent.

Co-occurring and current dependence upon opioid analgesics (n = 3) and NOA (n = 2), was seen; substances with prior co-occurring dependence also comprised nicotine (n = 8) and alcohol (n = 2) (Table 1).

3.2.2 Mono-dependence

Four patients reported a mono-dependence upon ZD: one was moderate and three were mild (lifetime and 12-month prevalence: 1%). The median duration was 2 years (Table 2, Figure 2).

3.2.3 De novo-dependence

Rates/prevalence were identical to the mono-dependence condition.

3.3 GPT dependence 3.3.1 Raw dependence

Eleven patients were currently and previously dependent upon GPT (gabapentin, n = 5; pregabalin, n = 4; both, n = 2) resulting in a lifetime prevalence of 2.75% (Table 2, Figure 2). Ten patients from this group remained dependent upon GPT at the time of the interview (12-month prevalence 2.5%) as well as upon NSAIDs; one person (previously dependent upon gabapentin) reported remission of GPT dependence; he was also previously dependent upon both oxycodone and tramadol (Figure 1). One patient from this group exhibited severe dependence (Table 2); he was serially dependent first upon gabapentin and then pregabalin. The remainder exhibited mild (n = 7) or moderate (n = 3) dependence.

3.3.2 Mono-dependence

Mono-dependence upon GPT was not found (Table 2, Figure 2).

3.3.3 De novo-dependence

Two cases (18.8% of all GPT-dependents; lifetime and 12-month prevalence: 0.5%) were identified whose first substance dependence comprised GPT; both of these patients, however, were simultaneously also dependent upon other NOA. At the time of the interview, both GPT-dependences had persisted for 2 and 15 years, and were rated as mild and moderate (Table 2, Figure 2).

3.4 Comparisons between the groups 3.4.1 Comparison of primary drug dependence prevalence between the groups

There were no significant differences in lifetime or 12-month prevalence rates of dependence, mono- or de novo dependence between these drug classes (χ2 = 0–0.76, p = 0.7–1). The GPT-mono-dependence was absent in our population. This was in contrast to BDZ- and ZD-mono-dependences (lifetime/12-month prevalence-rates: BDZ-mono-dependence 2.25%/0.75%, see Section 3.1.2., and ZD-mono-dependence 1%/1%, see Section 3.2.2).

3.4.2 Comparison of co-occurring substance dependence between the groups

Patients with BDZ dependence (n = 28) most commonly exhibited current or previous dependence upon opioid analgesics (n = 10; 36.7%), followed by alcohol and nicotine (n = 9 each; 32.1%) (Table 1).

Patients with ZD dependence (n = 17) most commonly exhibited current or previous dependence upon nicotine (n = 8; 47.1%), followed by opioids (n = 3; 17.6%) and then alcohol and NOA (n = 2 each; 11.8%) (Table 1).

Patients with GPT dependence (n = 11) most commonly exhibited current or previous dependence upon other NOA (10; 90.9%) with one patient (9.1%) exhibiting previous opioid dependence (Table 1).

Among opioid-dependent patients (n = 43, 10.75%), comorbid BDZ dependence was found in 10 patients (23.3%), ZD dependence in two patients (4.7%) and GPT in one patient (2.3%; p < 0.01).

4 DISCUSSION

There is growing concern over the misuse/abuse potential, dependence upon and addiction to, and harms attributable to GABA-mimetic drugs, which are prescribed increasingly throughout the developed world (Bonnet & Scherbaum, 2017; Brandt & Leong, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021). In 2018, over 56 million GPT prescriptions, over 47 million BDZ prescriptions and over 14 million ZD prescriptions were dispensed in the United States alone (MEPS, 2019).

Misuse potential of GABAmimetics has been proposed as being greatest for pregabalin, followed by gabapentin and then the long-acting BDZ clonazepam (Schifano et al., 2011), based upon self-reports posted on European websites by anonymous recreational drug users. In the present study, however, we found low BDZ or ZD-dependence rates without evidence of lone (“mono”) GPT-abuse/dependence. These results echo those seen among non-drug-abusing healthy subjects taking pregabalin in controlled laboratory settings, who did not experience the desire to take the drug again (Zacny et al., 2012).

In this first comparative clinical study about the dependence liabilities of GABAmimetics none of our sample met criteria for substance abuse (neglect of major roles; hazardous use; legal problems; use despite social/interpersonal problems), whether GABAmimetic or illicit drug. Our sample however, while possibly representative of a very broad swath of the global population certainly is not representative of all, and it should be pointed out that while many of these patients experienced comorbid dependence to alcohol, nicotine and other prescription drugs including opioid analgesics, no one among this sample had ever abused illicit drugs. This illicit drug-naivete might be the most relevant contextual factor for the low prevalence-rates of GABAmimetic mono-dependences in our population, in clear contrast to the known high misuse reports of BDZ and GPT among illicit substance-using persons (Abrahamsson et al., 2017; Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; Specka et al., 2011; Soyka M., 2017).

Along those lines, we found no cases of GPT-dependence (“mono-dependence”) which did not involve dependence upon other substances. Otherwise, the prevalence rates of BDZ- and ZD-mono-dependence, which are considered to correlate well with its (substance-specific) addictive power (Bonnet & Scherbaum, 2017; Bonnet et al., 2019), were low (lifetime prevalence 2.25% and 1% for BDZ and ZD, respectively). The lifetime prevalence rates of GABAmimetic dependence occurring with other comorbid substance dependence were higher for BDZ (7%) and ZD (4.25%) than for GPT (2.75%). In this context, we found the other GABAmimetics more often related to opioid dependence than were GPTs. Among opioid-dependent patients (n = 43, 10.75%), comorbid BDZ dependence was found in 10 patients (23.3%), ZD dependence in two patients (4.7%) and GPT dependence in one patient (2.3%; p < 0.01).

Even if social desirability and recall biases in the face-to-face SKID-I interview process cannot be totally excluded, the interpretation of the present study strengthens the likelihood of strong contextual factors influencing/modifying the growing series of GPT-misuse reports among opioid or polydrug users (Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; Peles et al., 2020). Many of these reports point to at least four major factors (“needs”) that can predispose to GPT-misuse among opioid addicts: (i) self-medication of withdrawal states, (ii) mitigation of opioid-induced hyperalgesia, (iii) self-medication of co-morbid disorders, and (iv) dose-dependent euphoria and relaxation, which might culminate in coma and respiratory depression/arrest – especially if combined with opioids and/or other CNS-depressants and/or relevant respiratory disease conditions (Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; McAnally et al., 2020; Peles et al., 2020). BDZ and ZD (as well as alcohol) share similar (and generally more potent) hedonistic/relaxing intoxication properties and overdose liabilities with GPT (Abrahamsson et al., 2017; Bonnet & Scherbaum, 2017, 2018; Crestani et al., 2000; Mateu-Gelabert et al., 2017; Soyka M., 2017). However, it is still unclear whether the aforementioned contextual factors (i) and (ii) are more specific for GPTs than for BDZ. BDZ might blunt the negative emotional responses to these conditions (as well as mitigating other comorbid conditions, i.e., contextual factor [iii]) in general (Mateu-Gelabert et al., 2017; May et al., 2020; Specka et al., 2011) as opposed to GPTs, which have been shown to be particularly effective in the treatment of opioid withdrawal syndrome (evidence level II to Ib) (Ahmed et al., 2019; Krupitsky, 2020) and approved for the treatment of neuropathic pain, where hyperalgesia central sensitization seem to play a key role for chronification (Dias et al., 2015). There likely exist further contextual factors (sociodemographic, psychiatric comorbidity presence including antecedent SUD, chronic pain conditions, etc.) that confound these outcomes (Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; Fonseca et al., 2021).

Gabapentinoids misuse and abuse may indeed be increasing, particularly in countries working to curb a prescription opioid epidemic by supply reduction (Skolnick, 2018). Gabapentinoids are widely distributed, easy to obtain, inexpensive, and better tolerated than BDZ, ZD or alcohol, with a larger therapeutic window than the latter substances in terms of fulfilling the increasing “needs” of an addicted population losing their prescription opioid source. Small wonder, then, that public concern regarding GPT continues to grow (Bonnet & Scherbaum, 2017; Evoy, Sadrameli, Contreras, Covvey, Peckham & Morrison, 2021; Smith et al., 2015). However, these concerns should pale in comparison to BDZ, ZD and alcohol, the latter of which continues to comprise the most prevalent, addictive and toxic GABA agonist available worldwide.

4.1 Limitations and strengths

From the standpoint of generalizability, nearly one fifth of the overall hospitalized population, selected randomly (Figure 1) participated in the original study providing reasonable representation of an elderly European hospital population; having said that Germany is not at present experiencing an opioid epidemic, and thus contextual factors as discussed above may render the results less generalizable to nations that are experiencing more widespread drug problems (Bonnet et al., 2021; Evoy, Sadrameli, Contreras, Covvey, & Peckham, 2021). A good portion of patients (1/3) refused to participate which diminishes the experimental value of a random sample and may introduce a non-response bias. Our sample displayed considerable female preponderance, which, however, is not atypical for a random geriatric sample.

From a methodologic standpoint, we relied exclusively on participants' self-reports which bears the risk of underreporting and interview biases. Although we used SKID-I, an instrument delivering the highest standard in the area of clinical interview diagnostics in Germany (Braehler et al., 2002; Strauβ & Schumacher, 2005), the interview method is limited to evaluation of the main problematic drug especially in situations of polydrug dependence. Objective measures such as blood and/or urine toxicologic data were not captured in the original sample. Those measures might have strengthened the validity of our results (especially in the analysis of mono-dependence.)

Despite the original dataset being collected nearly a decade ago (2013) already we were seeing considerable escalation of GPT prescriptions in Germany (at that time estimated at 80 million pregabalin and 40 million gabapentin defined daily doses, DDD) paralleling that seen in the U.S. and U.K. Conversely, we were already witnessing a trend of decreasing prescriptions for BDZ and other hypnotics including ZD, estimated around 200 million DDD at the time (Bonnet & Scherbaum, 2018; Fritze et al., 2018).

The study did not determine somatic or psychiatric diagnoses aside from substance dependence and/or addiction, and it is possible that some dependence symptoms (e.g., tolerance, withdrawal, drug-seeking behavior) resulted from a “pseudo-dependence” on an analgesic drug resulting from suboptimal pain treatment (Bonnet et al., 2019) and not from a true dependence. This could have resulted in overestimation of the dependence prevalence reported in the present study.

Nonetheless, we provide an index investigation into the comparative addictive power of BDZ, ZD, and GPT in a human sample which was the main strength of this study.

4.2 Future directions

These findings warrant an updated investigation given the ongoing worldwide increases in GABA-mimetic and especially GPT prescriptions. Stratification of study populations by prescription indication (especially given the increasing off-label use of the drugs) as well as major somatic and psychiatric comorbidities may also help shed light on contextual factors influencing potential misuse/abuse and the development of dependence on these ubiquitous agents.

5 CONCLUSION

In this elderly and illicit substance-naïve German general hospital patient sample, GPT were found to be less (if at all) addictive than BDZ and ZD by analysis of both overall prevalence but more importantly mono- and pure de novo-dependence. The severity of dependence was mostly moderate-to-mild for all substances. Benzodiazepines were significantly more frequently related to an opioid analgesic-dependence than were GPT. Nonetheless, the prescription of reinforcing/rewarding CNS-depressants, regardless of whether BDZ, ZD or GPT, requires very careful and individual benefit: risk stratification, along with effectiveness, harms and adherence monitoring, and vigilant tapering/discontinuation should be carried out when risks begin to outweigh benefits (McAnally et al., 2020).

ACKNOWLEDGMENT

The authors thanks Johanna-Cristina Strasser (née Coβmann) for having conducted the original face-to-face SKID-interviews, which data were re-evaluated here.

Open access funding enabled and organized by Projekt DEAL.

AUTHOR CONTRIBUTION

Study conception, analysis, data interpretation, drafting the article by both authors.

留言 (0)

沒有登入
gif