Prescriptions of psychotropic and somatic medications among patients with severe mental disorders and healthy controls in a naturalistic study

Abstract

Purpose

Psychotropic and somatic medications are both used in treating severe mental disorders (SMDs). Realistic estimates of the prevalence of use across medication categories are needed. We obtained this in a clinical cohort of patients with SMD and healthy controls (HCs).

Materials and methods

Prescriptions filled at Norwegian pharmacies the year before and after admittance to the Thematically Organized Psychosis (TOP) study were examined in 1406 patients with SMD (mean age 32.5 years, 48.2% women) and 920 HC (34.1 years, 46.2% women). Using data from the Norwegian Prescription Database (NorPD), the number of users in different anatomical therapeutic chemical (ATC) categories was compared using logistic regression. Population estimates were used as reference data.

Results

Use of antipsychotics (N05A), antiepileptics (N03A), antidepressants (N06A), anxiolytics (N05B), hypnotics and sedatives (N05C), anticholinergics (N04A), psychostimulants, attention deficit hyperactivity disorder and nootropic agents (N06B) and drugs for addiction disorders (N07B) was significantly more prevalent in patients with SMD than HC. Use of diabetes treatment (A10), antithrombotic drugs (B01), beta blockers (C07), lipid modifiers (C10), and thyroid and endocrine therapeutics (H03) was also more prevalent in patients with SMD, but with two exceptions somatic medication use was comparable to the general population. Among HC, there was low prevalence of use for most medication categories.

Conclusion

Patients were using psychiatric medications, but also several types of somatic medications, more often than HC. Still, somatic medication use was mostly not higher than in the general population. The results indicate that HC had low use of most medication types.

Keywords:

• Schizophrenia
• bipolar disorders
• psychotic disorders
• prescription drugs
• health register

Introduction

Psychotic disorders and bipolar disorders are considered severe mental disorders (SMDs) since they are associated with a substantial risk of functional impairment, chronic or relapsing/remitting illness courses, co-morbidities, and increased mortality [1–3]. Together, schizophrenia, other psychotic disorders, bipolar disorders, and severe unipolar depression affect around 3% of the population [4]. Patients with SMDs are at higher risk of non-communicable diseases such as diabetes, respiratory illness, and cardiovascular disease [5–7]. Respiratory and infectious diseases and metabolic syndrome were reported to have prevalence rates from 40 to 70% in patients with schizophrenia and 20–30% in patients with bipolar disorders [8–10].

Historically, the pharmacological treatment of SMDs was revolutionized in the 1950s with the availability of new and more effective medicines, such as antipsychotics, tricyclic antidepressants, and mood stabilizers. Since then, these psychotropic medication classes remain cornerstones in treating SMD [11]. Given the high prevalence of somatic comorbidities, patients with SMD also often use somatic medications. A Dutch study reported that more than 80% of hospitalized psychiatric patients used somatic medications for some reasons [12]. However, other studies have indicated that somatic comorbidities are under-detected [13,14] and treatment for these is under-utilized in this patient group [15]. Thus, obtaining realistic prevalence estimates of medication use for somatic comorbidities is of interest.

The procedure for selecting participants to healthy control (HC) groups in psychiatric research is an important issue. Shtasel et al. found that out of 1670 persons responding to a newspaper advertisement, less than 10% were deemed suitable as control subjects after telephone screening and in-person evaluation by a medical doctor [16]. Particularly in case-control studies, the control group is important since it provides the reference comparison, but HC are still rarely examined as thoroughly as the patients. This may lead to bias and influence the quality of research [17]. Examining the real-life use of psychotropic medication among HC is one way of determining the reliability of the screening and the validity of the HC group as a reference.

The aim of the current study was to investigate the prevalence of medication use for drugs acting on the nervous system and somatic medications in a large sample of patients diagnosed with SMDs according to a standardized research protocol and HC participants randomly drawn from the population register and screened in the same study. We determined the prevalence of medication use through linkage with a national prescription register, providing highly reliable data for prescriptions filled at pharmacies. We compared the prevalence of medication use between patients with SMD and HC, and we additionally obtained age-weighted population estimates as reference data. We hypothesized that the prevalence of use of psychiatric medications as well as medication classes used to treat metabolic syndrome is higher in patients with SMD compared with HCs and the general population. Furthermore, we hypothesized that due to the screening procedure, the HCs would use less medications, in particular psychiatric medications, than the general population.

Materials and methods

Participants

The Thematically Organized Psychosis (TOP) study is an ongoing naturalistic study conducted by the Norwegian Centre for Mental Disorders Research (NORMENT) in Oslo, Norway. Patients are recruited from psychiatric departments of five major hospitals in the catchment areas of Oslo. The inclusion criteria include age between 18 and 65 years, ability to give written consent, a diagnosis of non-organic psychosis or bipolar disorder according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) based on the Structural Clinical Interview for DSM-IV, Axis I Disorders (SCID-I) [18]. The exclusion criteria include any history of moderate or severe head injury, any neurological disorder, any autoimmune disease, and mental retardation (defined as IQ < 70).

Healthy controls were asked to participate in the study after having been randomly drawn from the national population register within the same geographical region as the patients. They were screened using unstructured interviews and the Primary Care Evaluation of Mental Disorders (PRIME-MD) [19], which is a screening tool designed to assess symptoms of common mental disorders. Healthy controls were excluded if they had a current psychiatric disorder, if they or any first-degree relatives had experienced an SMD, or if they had a history of substance abuse or dependence.

The patients were interviewed by trained clinical psychologists or physicians. As part of the assessment protocol, the patients also underwent a medical exam performed by a physician. This included interviewing the patients about their medical history and the presence of somatic disorders or conditions was recorded using a standardized form.

All participants had given written consent to use their information from the Norwegian Prescription Database (NorPD). The TOP study was approved by the Regional Committees for Medical and Health Research Ethics (REC South-Eastern Norway), approval 2009/2485-91 and this research work got adjoint ethical approval.

The Norwegian Prescription Register

The Norwegian Prescription Register is a centralized database where information from prescriptions filled (i.e. the medication is dispensed) at pharmacies is collected. Since 1 January 2004, all pharmacies in Norway send this information to NorPD. Patients admitted in hospitals may receive medicines directly without getting registered in NorPD [20]. The regulations for the NorPD require that data should be non-identifiable. Therefore, the merged data sets were anonymized.

Prescription data

The NorPD data were extracted based on anatomical therapeutic chemical (ATC) codes of drugs between 2004 and 2017. These ATC codes were described by the World Health Organisation’s Collaborating Centre for Drug Statistics Methodology (WHOCCDSM) [21]. To study the patterns of prescription, the NorPD data were categorized into drugs acting on the nervous system (ATC-N) and drugs acting on other organ systems (non-ATC-N), hereafter referred to as somatic medications. The first category includes all registered psychotropic medications, and we extracted data for 16 ATC-N codes at the third level of the ATC classification. For the second category, we extracted data for 23 non-ATC-N codes at the second level of the ATC classification. For further details, see Tables 1–4.

Table 1. Prescriptions of drugs acting on the central nervous system (ATC-N) the year before inclusion in TOP study.

		Prevalence					Test of difference^b
ATC-N	Therapeutic indications	Patients		Healthy controls		Population estimate^a, %	B	p	OR	95% CI
		n	%	N	%					Lower	Upper
N01A	General anesthetics	0	0.0	0	0.0	0.001	–	–	–	–	–
N01B	Local anesthetics	5	0.4	0	0.0	0.3	1.844^c	.114	6.321	0.700	834.640
N02A	Opioids	138	9.8	80	8.7	9.8	0.151	.306	1.163	0.871	1.555
N02B	Other analgesics and antipyretics	50	3.6	19	2.1	6.9	0.566	.038	1.762	1.030	3.012
N02C	Anti-migraine preparations	31	2.2	12	1.3	2.8	0.539	.117	1.713	0.873	3.362
N03A	Antiepileptics	318	22.6	2	0.2	2.1	4.719^c	<.0001	112.056	39.567	534.857
N04A	Anticholinergic agents	24	1.7	0	0.0	0.003	3.532^c	<.0001	34.192	4.759	4342.61
N04B	Dopaminergic agents	3	0.2	0	0.0	0.1	1.689^c	.177	5.414	0.525	727.781
N05A	Antipsychotics	931	66.2	2	0.2	2.5	6.567^c	<.0001	711.232	251.892	3391.41
N05B	Anxiolytics	282	20.1	14	1.5	3.3	2.808	<.0001	16.578	9.619	28.572
N05C	Hypnotics and sedatives	402	28.6	27	2.9	5.0	2.608	<.0001	13.578	9.091	20.278
N06A	Antidepressants	561	39.9	8	0.9	5.8	4.330	<.0001	75.931	37.538	153.593
N06B	Psychostimulants, agents used for ADHD and nootropics	27	1.9	0	0.0	1.2	3.531^c	<.0001	34.158	4.778	4332.93
N06D	Anti-dementia drugs	0	0.0	0	0.0	0.005	–	–	–	–	–
N07B	Drugs used in addictive disorders^d	18	1.3	3	0.3	0.6	1.353^c	.008	3.869	1.377	14.702
N07X	Other nervous system drugs	0	0.0	0	0.0	0.02	–	–	–	–	–

(–) Statistical test not applicable.

^a Weighted estimate as described in the ‘Materials and methods’ section.

^b Patients > healthy controls.

^c Statistical estimates calculated using Firth’s procedure.

^d Remedies for nicotine, alcohol and opioid addiction.

Table 2. Prescriptions of drugs acting on the central nervous system (ATC-N) the year after inclusion in TOP study.

		Prevalence					Test of difference^b
ATC-N	Therapeutic indications	Patients		Healthy controls		Population estimate^a, %	B	p	OR	95% CI
		n	%	N	%					Lower	Upper
N01A	General anesthetics	0	0.0	0	0.0	0.001	–	–	–	–	–
N01B	Local anesthetics	7	0.5	1	0.1	0.3	1.121^c	.164	3.067	0.664	29.224
N02A	Opioids	136	9.7	67	7.3	9.8	0.317	.043	1.373	1.011	1.864
N02B	Other analgesics and antipyretics	34	2.4	21	2.3	6.9	0.078	.781	1.082	0.623	1.877
N02C	Anti-migraine preparations	43	3.1	14	1.5	2.8	0.720	.021	2.054	1.114	3.789
N03A	Antiepileptics	447	31.8	5	0.5	2.1	4.501	<.0001	90.131	37.123	218.828
N04A	Anticholinergic agents	28	2.0	0	0.0	0.003	3.662^c	<.0001	38.939	5.468	4939.40
N04B	Dopaminergic agents	2	0.1	0	0.0	0.1	1.243^c	.372	3.465	0.270	482.510
N05A	Antipsychotics	1099	78.2	4	0.4	2.5	6.580^c	<.0001	720.539	318.621	2162.45
N05B	Anxiolytics	268	19.1	14	1.5	3.3	2.758	<.0001	15.772	9.142	27.211
N05C	Hypnotics and sedatives	401	28.5	25	2.7	5.0	2.691	<.0001	14.745	9.737	22.328
N06A	Antidepressants	536	38.1	21	2.3	5.8	3.284	<.0001	26.689	17.079	41.708
N06B	Psychostimulants, agents used for ADHD and nootropics	24	1.7	1	0.1	1.2	2.321^c	<.0001	10.185	2.638	91.469
N06D	Anti-dementia drugs	0	0.0	0	0.0	0.005	–	–	–	–	–
N07B	Drugs used in addictive disorders^d	23	1.6	3	0.3	0.6	1.533^c	.002	4.632	1.697	17.374
N07X	Other nervous system drugs	0	0.0	0	0.0	0.02	–	–	–	–	–

(–) Statistical test not applicable.

^a Weighted estimate as described in the ‘Materials and methods’ section.

^b Patients > healthy controls.

^c Statistical estimates calculated using Firth’s procedure.

^d Remedies for nicotine, alcohol and opioid addiction.

Table 3. Prescriptions of somatic medications (non-ATC-N) the year before inclusion in TOP study.

		Prevalence					Test of difference^b
Non-ATC-N	Therapeutic indications	Patients		Healthy controls		Population estimate^a, %	B	p	OR	95% CI
		N	%	N	%					Lower	Upper
A02	Remedies for acid-related disorders	58	4.1	28	3.0	6.2	0.350	.136	1.420	0.896	2.249
A03	Remedies for functional gastrointestinal disorders	25	1.8	7	0.8	1.2	0.805	.062	2.237	0.960	5.213
A05	Bile and liver therapy	0	0.0	1	0.1	0.05	−1.547^c	.307	0.213	0.001	4.100
A07	Antidiarrheal, intestinal antiseptics and anti-inflammatory drugs	13	0.9	4	0.1	1.2	0.643^c	.220	1.902	0.695	6.303
A08	Anti-obesity agents, excluding dietary supplements	12	0.9	6	0.7	0.1	0.243	.629	1.275	0.475	3.422
A10	Drugs for diabetes treatment	24	1.7	0	0.0	1.7	3.499^c	<.0001	33.082	4.604	4200.67
B01	Antithrombotic agents	21	1.5	5	0.5	2.1	1.071	.034	2.918	1.083	7.864
C01	Cardiac therapy	2	0.1	6	0.7	0.6	−1.283^c	.067	0.277	0.051	1.093
C02	Antihypertensives	0	0.0	0	0.0	0.07	–	–	–	–	–
C03	Diuretics	16	1.1	4	0.4	0.6	0.859^c	.087	2.360	0.890	7.706
C05	Vasoprotectors	16	1.1	16	1.7	1.2	−0.438	.221	0.645	0.320	1.301
C07	Beta blockers	22	1.6	5	0.5	1.8	1.120	.025	3.064	1.152	8.153
C08	Calcium antagonists	12	0.9	4	0.4	1.0	0.542^c	.316	1.7194	0.610	5.783
C09	Agents having an effect on the renin–angiotensin system	25	1.8	15	1.6	3.0	0.116	.729	1.123	0.581	2.170
C10	Lipid modifiers	25	1.8	5	0.5	2.4	1.259	.012	3.523	1.319	9.407
H02	Corticosteroids for systemic use	30	2.1	33	3.6	3.3	−0.530	.039	0.589	0.356	0.974
H03	Thyroid therapy	46	3.3	2	0.2	2.4	2.532^c	<.0001	12.578	4.246	61.191
L02	Endocrine therapy	0	0.0	3	0.3	0.3	−2.366^c	.046	0.0938	0.001	0.967
L03	Immuno-stimulants	0	0.0	0	0.0	0.08	–	–	–	–	–
L04	Immuno-suppressants	4	0.3	3	0.3	1.0	−0.123^c	.863	0.8842	0.215	3.963
M01	Anti-inflammatory and anti-rheumatic drugs	193	13.7	141	15.3	16.5	−0.121	.316	0.886	0.698	1.123
M02	Topical preparations for muscle and joint pain	8	0.6	11	1.2	1.1	−0.727^c	.108	0.4833	0.191	1.172
M03	Muscle relaxants	16	1.1	2	0.2	0.2	1.508^c	.009	4.5176	1.408	22.760
M04	Gout remedies	1	0.1	2	0.2	0.3	−0.814^c	.423	0.4430	0.040	3.357
M05	Agents for the treatment of bone diseases	3	0.2	1	0.1	0.1	0.492^c	.604	1.6355	0.269	16.877
R03	Remedies for obstructive pulmonary disease	84	6.0	43	4.7	5.8	0.277	.151	1.319	0.904	1.925

^a Weighted estimate as described in the ‘Materials and methods’ section.

^b Patients > healthy controls.

^c Statistical estimates calculated using Firth’s procedure.

Table 4. Prescriptions of somatic medications (non-ATC-N) the year after inclusion in TOP study.

		Prevalence					Test of difference^b
Non-ATC-N	Therapeutic indications	Patients		Healthy controls		Population estimate^a, %	B	p	OR	95% CI
		N	%	N	%					Lower	Upper
A02	Remedies for acid-related disorders	81	5.8	33	3.6	6.2	0.544	.010	1.723	1.137	2.611
A03	Remedies for functional gastrointestinal disorders	28	2.0	7	0.8	1.2	0.957	.024	2.605	1.131	6.000
A05	Bile and liver therapy	0	0.0	1	0.1	0.05	−1.438^c	.334	0.2374	0.002	4.380
A07	Antidiarrheal, intestinal antiseptics and anti-inflammatory drugs	14	1.0	4	0.4	1.2	0.777^c	.128	2.1749	0.810	7.142
A08	Anti-obesity agents, excluding dietary supplements	11	0.8	8	0.9	0.1	−0.127	.786	0.881	0.352	2.204
A10	Drugs for diabetes treatment	31	2.2	0	0.0	1.7	3.759^c	<.0001	42.905	6.050	5437.09
B01	Antithrombotic agents	30	2.1	8	0.9	2.1	0.927	.022	2.526	1.142	5.585
C01	Cardiac therapy	3	0.2	7	0.8	0.6	−1.159^c	.062	0.3137	0.076	1.058
C02	Antihypertensives	0	0.0	1	0.1	0.07	−1.440^c	.334	0.236	0.002	4.384
C03	Diuretics	19	1.4	3	0.3	0.6	1.277^c	.013	3.585	1.280	13.626
C05	Vasoprotectors	18	1.3	10	1.1	1.2	0.166	.678	1.180	0.540	2.577
C07	Beta blockers	31	2.2	9	1.0	1.8	0.884	.021	2.422	1.144	5.126
C08	Calcium antagonists	14	1.0	4	0.4	1.0	0.737^c	.159	2.089	0.761	6.945
C09	Agents having an effect on the renin–angiotensin system	28	2.0	16	1.7	3.0	0.156	.632	1.169	0.617	2.214
C10	Lipid modifiers	34	2.4	8	0.9	2.4	1.127	.005	3.086	1.403	6.787
H02	Corticosteroids for systemic use	21	1.5	33	3.6	3.3	−0.873	.002	0.418	0.240	0.728
H03	Thyroid therapy	54	3.8	3	0.3	2.4	2.370^c	<.0001	10.697	4.150	39.017
L02	Endocrine therapy	1	0.1	3	0.3	0.3	−1.303^c	.160	0.2717	0.026	1.662
L03	Immuno-stimulants	2	0.1	1	0.1	0.08	0.178^c	.863	1.1948	0.155	13.210
L04	Immuno-suppressants	4	0.3	2	0.2	1.0	0.231^c	.768	1.259	0.277	7.265
M01	Anti-inflammatory and anti-rheumatic drugs	193	13.7	136	14.8	16.5	−0.061	.618	0.941	0.741	1.195
M02	Topical preparations for muscle and joint pain	12	0.9	3	0.3	1.1	0.909^c	.107	2.481	0.831	9.699
M03	Muscle relaxants	9	0.6	2	0.2	0.2	0.955^c	.146	2.598	0.735	13.613
M04	Gout remedies	1	0.1	1	0.1	0.3	−0.305^c	.793	0.737	0.059	9.134
M05	Agents for the treatment of bone diseases	2	0.1	1	0.1	0.1	0.399^c	.726	1.4903	0.071	8.125
R03	Remedies for obstructive pulmonary disease	93	6.6	57	6.2	5.8	0.088	.612	1.092	0.776	1.537

^a Weighted estimate as described in the ‘Materials and methods’ section.

^b Patients > healthy controls.

^c Statistical estimates calculated using Firth’s procedure.

Investigation of medication use by both patients and control groups was done by merging this data from the NorPD register with TOP study data by the unique 11 digits personal number assigned to every resident in Norway. Medication use was investigated using a case-control cohort study approach. We considered medication use both one year before and one year after inclusion in the TOP study. The final sample with complete data comprised 1406 patients and 920 controls, yielding a total participant number of 2326.

Analyses

Statistical analyses were performed using SPSS version 25 (SPSS Inc., Chicago, IL). The significance threshold was set as p < .05 and data were presented both as frequency (number of users) and percentage (prevalence) within patients and HC. Crosstabulation was performed to display the use of ATC codes (ATC-N and non-ATC-N) between patient and control groups with one year time intervals before and after inclusion in the TOP study. To test if the prevalence of medication use was different between patients and control subjects, we performed logistic regression controlling for age and gender. Firth logistic regression was performed in cases where the number of users in any patient or control group was less than 5. Reference population data were obtained using aggregated data available from the NorPD (reseptregisteret.no). We extracted national prevalence estimates (use observed per 1000 person) from 2017 for five-year age bins ranging from 15 to 70 years, including both genders. To make the population-based prevalence estimates comparable to the study sample, we calculated an age-weighted prevalence estimate for each medication category, using the proportion of the study sample within each age bin as weights.

Results

Demographic data

Patients (n = 1406) and HC (n = 920) had similar gender distributions: 51.8% and 53.8% men and 48.2% and 46.2% women among patients and controls, respectively. The mean age (standard deviation) among patients and controls was 32.5 (10.8) and 34.1 (9.5) years, respectively.

Of the patients, 515 (36.6%) were diagnosed with schizophrenia, 51 (3.6%) with schizophreniform disorder, 130 (9.2%) with schizoaffective disorder, 296 (21.1%) with bipolar I disorder, 144 (10.2%) with bipolar II disorder, 30 (2.1%) with bipolar disorder not otherwise specified, 54 (3.8%) with major depressive disorder and 183 (13.0%) with other psychosis. For three patients, the diagnosis was missing. Of the patients, 524 (37.3%) had never experienced any psychiatric hospital admission, whereas 392 (27.9%) had been hospitalized once, 352 (25%) 2–5 times, 63 (4.5%) 6–10 times, and 27 (1.9%) 11 times or more. For 48 patients, data on hospital admissions were missing. The prevalence of a range of somatic disorders or conditions among patients is included in Supplementary Table 1.

Use of central nervous system medications (ATC-N)

Prescription frequencies for central nervous system medication (ATC-N) in patients with SMD, HC and age-weighted general populations estimates are shown in Tables 1 and 2. During the year before inclusion to the study, patients with SMD were significantly more often prescribed psychotropic drugs compared to HC. The distribution in patients with SMD vs. HC was: antiepileptics (N03A; 23% vs. 0.2%), anticholinergics (N04A; 1.7% vs. 0.0%), antipsychotics (N05A; 66% vs. 0.2%), anxiolytics (N05B; 20% vs. 1.5%), hypnotics and sedatives (N05C; 29% vs. 3%), antidepressants (N06A; 40% vs. 0.9%), psychostimulants, attention deficit hyperactivity disorder (ADHD) and nootropic agents (N06B; 1.9% vs. 0.0%), and drugs for addiction disorders (N06D; 1.3% vs. 0.3%) one year before inclusion to the study. The corresponding population estimates were 2.1% (N03A), 0.003% (N04A), 2.5% (N05A), 3.3% (N05B), 5.0% (N05C), 5.8% (N06A), 1.2% (N06B) and 0.6% (N07B). The prevalence of medications prescribed in both groups was mostly similar the year after recruitment in the TOP study compared to the year before recruitment, except for a higher percentage of patients using antipsychotics (N05A; 78%) and antiepileptics (N03A; 32%). General anesthetics (N01A), dementia treatments (N06D) and other drugs working on the nervous system (N07X) were not prescribed to any of the participants. These three treatment groups were rarely used in the general population with age-weighted estimates of 0.001%, 0.005% and 0.02%, respectively.

Use of somatic medications

Data on prescriptions of non-ATC-N drugs in patients with SMD, HC and age-weighted general populations estimates are shown in Tables 3 and 4. The most frequently prescribed categories during the year before study inclusion were anti-inflammatory and antirheumatic drugs (M01; 13.7% vs. 15.3%), remedies for obstructive pulmonary disease (R03; 6.0% vs. 4.7%), remedies for acid-related disorders (A02; 4.1% vs. 3.0%), corticosteroids for systemic use (H02; 2.1% vs. 3.6%) and thyroid therapeutics (H03; 3.3% vs. 0.2%) in patients with SMD and HC, respectively. The corresponding population estimates were 16.5% (M01), 5.8% (R03), 6.2% (A02), 3.3% (H02), and 2.4% (H03). None of the patients were prescribed bile and liver therapeutics (A05) and the specific antihypertensives in ATC group C02, whereas none of the controls were prescribed drugs for diabetes treatment (A10). Population estimates for A05 and C02 were 0.05% and 0.07%, respectively, whereas drugs for diabetes treatment (A10) were used by 1.7% of the matched general population. Compared to HC, patients were significantly more often prescribed drugs for diabetes treatment (A10; 1.7% vs. 0%), antithrombotic agents (B01; 1.5% vs. 0.5%), beta blockers (C07; 1.6% vs. 0.5%), lipid modifiers (C10; 1.8% vs. 0.5%), thyroid therapeutics (H03; 3.3% vs. 0.2%) and muscle relaxants (M03; 1.1% vs. 0.2%) during the year before recruitment, whereas controls were prescribed corticosteroids for systemic use more often than patients (H02; 3.6% vs. 2.1%). Populations estimates for B01, C07 and C10 were 2.1%, 1.8% and 2.4%, respectively, thus nominally similar to or higher than the prescription frequencies for both patients with SMD and HC. Population estimates for drugs for diabetes treatment (A10) were lower or similar to patients with SMD, whereas for thyroid therapeutics (H03) and corticosteroids for systemic use (H02) the prevalence estimates fell between patients with SMD and HC. Similar frequency distributions were seen the year after recruitment but with significantly higher prescription rates in patients of drugs for acid-related disorders (A02; 5.8% vs. 3.6%) and functional gastrointestinal disorders (A03; 2.0% vs. 0.8%), whereas the difference between patients and controls for muscle relaxants was no longer statistically significant.

Discussion

In this study, we present prescription registry-based prevalence estimates for medication use in patients with SMD and screened HC, alongside with age-weighted population estimates. The prevalence of use was assessed for a wide range of medication types. Thus, the presented data provide rich information that may be useful to researchers, public health workers, and clinicians who wish to obtain realistic estimates of medication use in this patient group.

Patients with SMD are at higher risk of metabolic abnormalities [22]. These may contribute to an increased disease risk and premature mortality [23,24]. Still, there are data indicating that medical risk and comorbidities are not yet sufficiently monitored and treated in this group [25,26]. In our sample, the use of lipid modifiers was significantly more prevalent in patients than HC; however, it was numerically identical to the prevalence of use in the general population (2.4% treated). A recent Danish registry study focusing on patients with schizophrenia found that routine screening for dyslipidemia increased from 2000 to 2012, following improved guidelines, but it remained suboptimal: close to a third of patients were not monitored during the three months before and after receiving the diagnosis at the end of the study period [26]. Diabetes treatment was more prevalent in patients compared with HC in our study, with a numerically higher prevalence estimate than the general population (2.2% vs. 1.7%, respectively). A previous registry-based study found a threefold increased risk of diabetes in patients with schizophrenia compared to healthy individuals [27]. Prescriptions for thyroid abnormalities were more common in patients compared to HC in our sample, with higher prevalence also found in patients compared with the general population (3.8% vs. 2.4%, respectively). This is in line with our previous study showing thyroid abnormalities related to having a diagnosis of an SMD and the use of antipsychotic treatment [28]. Our data suggest that known hormonal disorders among patients were not undertreated, given that the prevalence of these disorders (including both diabetes and thyroid disease) was found to be 5.7%. However, there is still a possibility of undetected cases.

Among the non-psychiatric central nervous system medications, more patients used opioids, other analgesics and antipyretics, and anti-migraine preparations than controls during the year before or after inclusion in the study. However, the frequency of opioid and anti-migraine use in patients was overall similar to the population estimates. In contrast, the use of other analgesics and antipyretics were lower than the population estimate (6.9%) in both the patient (up to 3.6%) and control groups (up to 2.3%). Given the prevalence of muscle- or joint-related pain conditions (2.6%) and migraine or other types of headaches (2.5%) among patients, our data do not indicate over-prescription of these medication types. From meta-analyses, it has been reported that in experimental studies patients with schizophrenia had a decreased pain sensitivity [29], whereas in clinical studies similar levels of pain were reported as among age- and gender-matched controls [30]. The reason for this discrepancy between experimental and clinical studies is unknown, but may stem from underreporting, higher pain thresholds or lower help seeking behaviors among patients with schizophrenia [30]. In contrast, patients with bipolar disorder had increased levels of pain [31]. Among Danish users of selective serotonin reuptake inhibitors, medications used to treat depression and anxiety, 8% used paracetamol, one of the compounds of the ATC-group N02B, concomitantly [32]. The lower prescription frequency among controls in the present study may be a result of the screening procedure, or possibly that the controls are more likely to buy the pain relievers without a prescription. Other analgesics and antipyretics are available prescription free in Norway. The prescription register data do not include information about over-the-counter medication use.

Overall, the HC group in the present study had low prescription rates of any kind of medications, usually of about the same or lower level than the general population. This may be anticipated, given that the controls had been screened to be healthy. Still, some of the controls had been prescribed medications with indications for mental and other disorders, indicating that the control group was not completely healthy. The reasons for this are unclear but may include difficulties disclosing information during the interaction with a research team [33] or simply forgetfulness [34], for example in cases where the individual was on continued maintenance treatment for a psychiatric disorder in remission. Procedures for recruitment of control participants are debated and inclusion criteria vary between studies. Should controls be screened to make the difference between the case and comparison groups as distinct as possible, enhancing the likelihood of detecting illness-related differences, or is it desirable to have a control group mirroring the general population to improve representativeness? The present study recruited the HC participants based on a registry of the general population which increased representativeness, but also performed screening for psychiatric disorders and other illnesses and disorders believed to influence the nervous system. However, the screening procedure was not as extensive as the patient investigation, which has been advocated by some researchers [35] in order to ensure the validity of the control group [17].

As expected [36,37], the prevalence of psychiatric medication use was substantially higher among patients compared with HC the year before and after recruitment in the TOP study. Higher use of antipsychotics, antiepileptics, antidepressants, anxiolytics, hypnotics and sedatives, anticholinergics, psychostimulants, ADHD and nootropic agents and drugs for addiction disorders was observed for patients with SMD. For somatic medication types, patients had significantly higher use of diabetes treatment, antithrombotic drugs, beta blockers, lipid modifiers, and thyroid and endocrine therapeutics. However, it is worth noting that the difference was relatively modest and, importantly, for most of these categories, patients appeared to have a similar or lower prevalence of use than the general population. One medication category, corticosteroids for systemic use, was more often prescribed to HC than patients.

Strengths and limitations

Strengths of our study include the relatively large sample of patients and controls characterized through a standardized research protocol and the use of highly reliable registry data from NorPD. One limitation is that the prescription registry only covers prescriptions filled at pharmacies, whereas medication dispensed during hospitalization is not covered. However, according to Statistics Norway, the average duration of psychiatric admissions has been shorter in the past decades, with an average of 20 days in 2017 [38]. This reduces the influence of bias resulting from medication not being recorded during hospital stays. Another limitation is that some of the medications can be obtained without a prescription in Norway and are thus not registered in NorPD. This may cast doubts on the results, especially regarding analgesics and antipyretics; however, patients and HC have equal access to over-the-counter medications. Furthermore, our data do not consider the medication obtained at visits abroad, although this is likely to be a minor cause of uncertainty. Lastly, while we know that the medication has been delivered to the patient, we do not know if the patient eventually took the drugs as intended. The present study has this limitation in common with in practice all research based on pharmacological registers.

Conclusion

Patients with SMD had a relatively low prevalence of use for most categories of somatic medication. While significant differences vis-à-vis the HC sample were observed for specific medication types, the medication use among patients generally appeared similar to the population reference data. Healthy control participants had low use of most medication types and reference data pointed toward higher medication use in the population. Despite observing minimal use of some categories of psychotropic medication, the overall low prevalence suggests that the screening of HC had been effective.

Disclosure statement

OAA has received speaker’s honorarium from Lundbeck, Janssen and Sunovion and is a consultant for Cortechs.ai. The other authors report no conflicts of interest.

Additional information

Funding

The work was supported by The Research Council of Norway (Grant Numbers 223273, 274359, and 300309) and the South-Eastern Norway Regional Health Authority (Grant Numbers 2019-108 and 2022-073).

Notes on contributors

Dur E. Shahnaz Shafi

Dur E. Shahnaz Shafi, MSc, is currently a pharmacist and working as a pharmacy manager at Vitusapotek, Norway.

Kjetil Nordbø Jørgensen

Kjetil Nordbø Jørgensen, cand. psychol, PhD, is currently a Researcher at the Centre for Precision Psychiatry, Institute of Clinical Medicine, University of Oslo, Norway, and a specialist in clinical psychology at Drammen DPS, Vestre Viken Hospital, Drammen, Norway.

Thomas Bjella

Thomas Bjella, MSc, is currently a Research coordinator at the Division of Mental Health and Addiction, Oslo University Hospital and Institute of Clinical Medicine, University of Oslo, Norway.

Ragnar Nesvåg

Ragnar Nesvåg, MD, PhD, is currently a department director at Norwegian Institute of Public Health in Oslo, and part-time consultant psychiatrist at Oslo University Hospital.

Ingrid Dieset

Ingrid Dieset, MD, PhD, is currently an Associate Professor at the Institute of Clinical Medicine, University of Oslo, Norway, and a specialist in psychiatry at Oslo University Hospital, Oslo, Norway.

Ingrid Melle

Ingrid Melle, MD, PhD, is currently Professor at the Adult Psychiatry Department, Institute of Clinical Medicine, University of Oslo and the Clinical Psychosis Research Section, Oslo University Hospital.

Ole A. Andreassen

Ole A. Andreassen, MD, PhD, is currently Professor in psychiatry, director of the NORMENT Centre, focusing on translational research from disease mechanisms to clinical impact. Has contributed to better understanding of causal factors of severe mental disorders, including immune factors, and recently focusing on large Nordic biobank and registry data for developing tools for precision psychiatry.

Erik G. Jönsson

Erik G. Jönsson, MD, PhD, is currently Professor Emeritus at the Institute of Clinical Medicine, University of Oslo, Norway and active at the Centre for Psychiatry Research, Department of Clinical Neuroscience & Stockholm Health Care Services, Stockholm Region, Stockholm, Sweden.