Chlordiazepoxide.

The full risk profile is substantial and the modern data are less flattering than the historical reputation. Dependence and withdrawal are established: a review of the benzodiazepine withdrawal syndrome documents reactions with chlordiazepoxide among other agents, notes that risk rises with dose and duration, and records that withdrawal can occur even at recommended doses after short courses ^[19]. In a controlled trial of sudden discontinuation after long term anxiolytic treatment, patients withdrawn abruptly showed a clear symptomatic response ^[20], and in mice, 21 days of escalating chlordiazepoxide followed by cessation produced a significant anxiogenic profile on the elevated plus maze at 24 hours ^[21]. Sedation and respiratory depression are the acute hazards, and they are quantified. Among 113 patients with 133 intensive care admissions after receiving at least 200 mg of chlordiazepoxide for alcohol withdrawal, the commonest reasons for admission were respiratory insufficiency at 39 percent and somnolence at 39 percent, mechanical ventilation was used in 71 percent of admissions, 36 percent were followed by hospital readmission within 90 days, and 90 day mortality in the over sedated group was 21 percent ^[11]. A prospective pharmacokinetic study in the same setting found that the correlation between cumulative dose and plasma concentration was poor, so dosing cannot be titrated reliably by concentration and therapeutic drug monitoring is not usable to guide it ^[12]. A case report describes a patient requiring life support from accumulated active metabolites after treatment for alcohol abstinence ^[14], and active benzodiazepine metabolites were still present in urine during the third and fourth weeks after drinking stopped in patients given chlordiazepoxide during withdrawal, at concentrations as high as or higher than after therapeutic dosing ^[15]. Hepatic impairment amplifies all of this; because unmetabolised chlordiazepoxide has little activity, delayed biotransformation can lead to dose stacking and then a delayed, profound and prolonged sedative effect once metabolites finally form ^[13]. Cimetidine substantially prolongs its elimination ^[23], and alcohol, opioids and other central nervous system depressants are additive and dangerous. Disinhibited hostility is a real and specific effect of this molecule rather than a class generality. In a controlled small group setting, chlordiazepoxide increased hostility more than placebo ^[16], a second study found chlordiazepoxide induced increases in motivational hostility more frequent and intense than with placebo or oxazepam ^[17], and the follow up on verbal behaviour reproduced the finding for interpersonal hostility after frustration ^[18]. Cognitive and psychomotor effects are consistent: chlordiazepoxide lactam impaired immediate memory and was the only agent in a two week double blind study to impair associative learning on its own, with alcohol worsening every arm ^[25], and reviews of Valium and Librium document psychomotor slowing and hangover effects ^[26]. In elderly people the risk is injury; higher doses of chlordiazepoxide, oxazepam and flurazepam carried the greatest injury risk in a five year prospective cohort of 253,244 new elderly users ^[27]. Its very long half life cuts both ways: withdrawal is gentler on discontinuation, but washout is slow when adverse effects or pregnancy make rapid clearance desirable ^[28]. Abuse liability is real but moderate within the class; chlordiazepoxide maintained self injection above vehicle control in baboons though at generally submaximal rates and below triazolam ^[29], and it only partially shared discriminative effects with midazolam with a low rate of self administration ^[30]. Pregnancy data are mixed but reassuring at therapeutic doses. A Hungarian population based case control study of 22,865 malformation cases and 38,151 controls found a raised odds ratio of 2.5 for congenital cardiovascular malformations with second and third month exposure, but the exposure data were mainly maternal self report and no specific malformation type was raised, so the authors concluded a substantial teratogenic risk is unlikely ^[31]. Thirty five women who self poisoned with 20 to 300 mg during pregnancy delivered live infants with no excess of congenital abnormality against sibling controls, though dose dependent intrauterine growth retardation was seen ^[32]. Four independent approaches in Hungary found no association between benzodiazepines and facial clefting ^[33], and a critical review concluded first trimester benzodiazepine exposure is not clearly associated with major malformations, naming diazepam and chlordiazepoxide as the better studied choices ^[34]. Genotoxicity and carcinogenicity data are limited and largely unremarkable; a review of the genetic toxicology of four common benzodiazepines covers chlordiazepoxide ^[36], and rats given chlordiazepoxide with sodium nitrite showed no tumour initiating effect in the liver ^[37].

Chlordiazepoxide is the origin point of the whole class. Leo Sternbach at Hoffmann-La Roche revisited a set of benzheptoxdiazine compounds he had worked on in the 1930s, and the molecule that became chlordiazepoxide emerged from a reaction whose product was not what the chemistry predicted; a compound set aside during a laboratory cleanup was submitted for screening almost as an afterthought and proved strongly active ^{[2] [1]}. Randall and colleagues published its pharmacology in 1960, describing taming, muscle relaxation and anticonvulsant activity under the name methaminodiazepoxide ^[3]. It was patented in 1958 and marketed as Librium in 1960, with diazepam following as Valium in 1963 ^[4]. Its introduction was a genuine break with the barbiturates, chiefly because it lacked their respiratory depression at ordinary doses, and benzodiazepines topped prescribing lists by the mid to late 1970s before dependence concerns reshaped the field in the 1980s ^[4]. Kaim's 1969 four drug comparison in acute alcohol withdrawal fixed the indication for which it is still used ^[5]. It was sold in eastern Europe and the Soviet Union as Elenium and Chlozepid, and remains available in combination products with clidinium and with amitriptyline.

Chlordiazepoxide carries the reputation of the founding benzodiazepine and, in the United Kingdom and much of Europe, of the default drug for medically supervised alcohol detoxification. Guidelines have long preferred it and diazepam over lorazepam for withdrawal, and a 2026 analysis argues that lorazepam's dominance in American practice traces to three ghostwritten comparison papers from 1983 and 1984 rather than to superior evidence ^[38]. Against that, the recent Danish intensive care literature has pushed hard in the other direction, describing dose stacking, unreliable titration and a high rate of mechanical ventilation ^{[11] [12]}. It also carries the odd distinction of being the benzodiazepine most clearly linked to disinhibited hostility in controlled human studies ^[16]. Its standing today is that of a historically important, still useful withdrawal agent whose kinetics demand caution in the liver impaired and the elderly.