Next Article in Journal
The Global Prevalence and Associated Factors of Burnout among Emergency Department Healthcare Workers and the Impact of the COVID-19 Pandemic: A Systematic Review and Meta-Analysis
Previous Article in Journal
Effects of Social Networks on Job Performance of Individuals among the Hypertension Management Teams in Rural China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Healthcare
Volume 11
Issue 15
10.3390/healthcare11152219
healthcare-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Baseline Profiles of Drug Prescriptions Prior to Diagnosis of Mild Cognitive Impairment (MCI) Obtained by Latent Class Analysis (LCA), and Assessment of Their Association with Conversion to Dementia

by
Carmen Gómez-Gómez
1,*,
Miguel Ángel Moya-Molina
2,
Manuel Jesús Tey-Aguilera
1,
Jorge Flores-Azofra
1 and
Juan Luis González-Caballero
3
1
Department of Biochemistry and Molecular Biology, School of Medicine, University of Cadiz, 11002 Cádiz, Spain
2
Department of Neurology, Hospital Universitario Puerta del Mar (HUPM), University of Cadiz, 11009 Cádiz, Spain
3
Department of Statistics and Operations Research, Faculty of Medicine, University of Cadiz, 11002 Cádiz, Spain
*
Author to whom correspondence should be addressed.
Healthcare 2023, 11(15), 2219; https://doi.org/10.3390/healthcare11152219
Submission received: 13 June 2023 / Revised: 27 July 2023 / Accepted: 31 July 2023 / Published: 7 August 2023
(This article belongs to the Section Chronic Care)
Browse Figures
Versions Notes

Abstract

:
Polypharmacy has been linked to cognitive decline. However, interventions targeting modifiable risk factors, some of which are targets of the most commonly used drugs, could reduce the prevalence of dementia. Our aim was to determine the drug prescription regimen at baseline, prior to the diagnosis of mild cognitive impairment (MCI), and its possible association with progression to dementia. Data were collected from the electronic medical records of 342 MCI outpatients diagnosed during 2006–2017 at their first neurology consultation. We followed the classical three-step method of statistical analysis, starting with a Latent Class Analysis (LCA) to discover subgroups of drug prescription probability. Half of the patients were under polypharmacy (≥5 drugs), 17.5% had no recorded medication, 33.3% progressed to dementia (94.7% in ≤5 years), and 84.1% of them to Alzheimer’s disease (AD). According to the LCA and based on 20 therapeutic indicators obtained from 240 substances and regrouped according the Anatomical Therapeutic Chemical Classification, we identified a four-profile model: (1) low (35.7% of patients); (2) mixed (28.7%); (3) cardio-metabolic (19.3%); and (4) psychotropic (16.4%). The binomial regression logistic model showed that profiles 2 and 3 (and 4 for AD), with a higher drug prescription conditioned probability against classic risk factors, were protective than profile 1 (OR = 0.421, p = 0.004; OR = 0.278, p = 0.000; OR = 0.457, p = 0.040, respectively), despite polypharmacy being significant in profiles 2 and 3 (mean > 7 drugs) vs. profile 1 (1.4 ± 1.6) (p = 0.000). Patients in the latter group were not significantly older, although being aged 65–79 years old quadrupled (OR = 4.217, p = 000) and being >79 tripled (OR = 2.945, p = 0.010) the conversion risk compared to patients <65 years old. According to the proposed analytical model, profiling the heterogeneous association of risk factors, which were taken prior to diagnosis, could be explored as an indicator of prior care and a predictor of conversion to dementia.

1. Introduction

Multiple chronic health conditions contribute to frailty in aging, with cognitive impairment being one of their manifestations. These conditions are often associated with polypharmacy [1,2,3,4]. Polypharmacy has been considered both a risk factor for and an indicator of cognitive status [5,6]. The potential protective or detrimental impacts on cognition of certain types of drugs has been analyzed individually for oral antidiabetics [7,8], antihypertensive and lipid-lowering agents [9], antidepressants [10], and benzodiazepines and related drugs [11]. It has been widely debated whether cognitive decline linked with polypharmacy could be attributed to the drugs used, drug–drug interactions, or the side effects of the medications [12,13]. Notably, some medications taken by older people have an anticholinergic effect (e.g., those taken for urinary frequency/incontinence, antidepressants, anti-psychotics, anti-parkinsonian, anti-arrhythmic drugs, antihistamines, bronchodilators, etc.) [14]. This has been posed as a potential explanation for the influence of a multidrug regimen on cognitive function [15].
Conversely, several studies have focused on mild cognitive impairment (MCI), a clinical entity representing a transitional state between normal cognition and dementia [16]. Not all patients with MCI necessarily deteriorate; some can remain stable or even revert to normal cognitive function over time. This depends on the diagnostic criteria, the clinical subtype [17], and the coexistence of modifiable midlife and later life risk factors, such as cardiovascular factors, cardiovascular and cerebrovascular diseases, psychiatric symptoms, and social conditions [18,19]. Moreover, early intervention targeting some of these factors has been shown to reduce the prevalence of dementia by 37% [20].
The aim of this study was to determine the baseline drug prescription regimens of MCI patients before their first neurological consultation. Given that these patients, who are typically older, are likely to be prescribed multiple drugs, Latent Class Analysis (LCA) was used to classify individuals into a limited number of multi-drug prescription subgroups. In light of the existing controversy, another objective was to assess the association between baseline medication profiles, potential polypharmacy, and the progression from MCI to a major neurocognitive disorder.

2. Materials and Methods

2.1. Data Source

This retrospective study was based on outpatients from the Department of Neurology at the Hospital Universitario Puerta del Mar (HUPM) (Cádiz, southern Spain). On arrival, neurologists examined each case according to a standardized protocol that includes clinical history, background co-morbidities, a general and neurological examination, a battery of cognitive and neuropsychiatric tests (MMSE (Mini-Mental State Examination), ADAS-cog (Alzheimer’s Disease Assessment Scale-Cognitive Subscale), ADL (Activities of Daily Living)), and blood sampling. Following this, the diagnosis of dementia is eminently clinical and based on established criteria [21,22,23,24]. MCI was diagnosed using modified Petersen criteria [25]: objective impairment in cognitive testing not severe enough to define dementia; normal general cognitive function; age- and education-corrected MMSE scored >23.8; absence of impairment in ADLs.
This information was stored in an electronic medical record containing the patient’s history from previous examinations carried out by their General Practitioner (GP) or other specialist physicians. From this data source, all patients diagnosed with MCI for the first time between February 2006 and March 2017 were consecutively analyzed, resulting in 342 patients fulfilling the inclusion criteria (see Appendix A Figure A1). The variables were recorded in an Excel database.

2.2. Baseline Characteristics

Sex, age, and family history of dementia (siblings or parents) were collected. Age was categorized into three intervals: <65, 65–79, and >79 years. Currently prescribed drugs were recorded, which were taken from a datasheet provided for that purpose or from the anamnesis of other physicians. Each active compound was identified by drug term (at the 5th level), then regrouped into therapeutic group (at the 2nd level), or subgroups (at the 3rd or 4th levels) when considered more appropriate, according to the Anatomical Therapeutic Chemical Classification (ATC code) (available at www.whocc.no/atc_ddd_index/last accessed on 1 September 2021). The number of medicines prescribed was also recorded. Polypharmacy was defined as the concomitant use of ≥5 active drugs, and excessive polypharmacy as ≥10 drugs [26]. This variable was then categorized into three groups: no polypharmacy (<5 drugs); moderate polypharmacy (5–9 drugs); and excessive polypharmacy (≥10 drugs). Anticholinergic drugs scoring 3 were obtained from http://www.acbcalc.com/ (last accessed on 30 September 2021), because of their categorization as increasing the risk of dementia/impaired cognitive performance [27].

2.3. Conversion

As part of their regular neurological care, the patients were followed-up every 6 months. We checked medical records for possible changes in diagnosis. A “conversion interval” variable was defined as the elapsed time from the registered date of MCI diagnosis to the date of the new diagnosis. A reasonable minimum of three years after the last online registration of the individual (March 2020) was used to examine possible changes [17] (see Appendix A Figure A1).

2.4. Statistical Analysis

The statistical analyses were carried out irrespective of who compiled the database and the neurologist who provided the anonymized data. Categorical variables are expressed as number and percentage of the observed data; numeric variables are represented as mean ± standard deviation.
Prior to LCA, the raw data were regrouped into 20 indicators obtained by regrouping 240 substances (see above) to better reflect the main pharmacological indications and underlying medical conditions of the patients [28]. First, we explored the solution with one class to obtain the initial fitting parameters. We increased the number of classes until we obtained the best fitting model, while respecting theoretical parsimony [29]. We used several criteria, such as the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), the Adjusted BIC (aBIC), and the Lo–Mendell–Rubin (LMR), and bootstrap likelihood ratio (BLRT) tests. Smaller values for all information criteria (reductions of more than 10 points) were considered improvements in the model fit. An entropy value between 0 and 1 allowed us to measure the uncertainty of the classification obtained, whereby values ≥0.80 indicated a strong separation and were used identify the different classes [30,31,32,33].
Second, the distribution of variables among latent class members was analyzed. Next, the characteristic of converted vs. non-converted patients were examined and, following the classical three-step method [33], and latent class membership was included as an independent categorical variable. The association between categorical variables was contrasted using a χ2 test, or if these conditions were not verified, with Fisher’s exact test. Comparisons of quantitative variables were performed using the Student’s t-test.
Third, a binomial logistic regression model was performed to assess the association of baseline drug prescription profiles and other potential contributing factors to the conversion from MCI to dementia, and specifically to AD. Predictors associated with the odds ratio (OR) were estimated with a 95% confidence interval (CI) using the Wald test, with p < 0.05 considered statistically significant. Data were processed using IBM SPSS Statistics v24, Epidat 3.1 software, and LCA was conducted using Mplus v8.4.

3. Results

Of the patients, 66.5% (n = 225) were female and 78.1% (267) were ≥65 years old (age 72.1 ± 9.5, range = 42–97 years).

3.1. Drug Prescription at Baseline

The mean number of drugs used was 4.8 ± 3.8 (range = 0–20). Individuals without a prescription accounted for 17.5% (n = 60) of the patients, and 49.7% (n = 170) were under polypharmacy, 22.4% (n = 38) of whom were under excessive polypharmacy. Polypharmacy was significantly higher in those aged ≥65 (53.4% of these patients) compared to those <65 years old (38.7%) (p = 0.007).
The most commonly used therapeutic subgroups were antihypertensive drugs (including C02AC, C03, C07, C08, and C09) (49.8% of patients, n = 166), drugs for acid-related disorders (A02) (42.2% n = 143), anxiolytics (N05B) (33%, n = 112), lipid-modifying agents (C10) (30.1%, n = 102), antithrombotic agents (B01) (29.2%, n = 99), antidepressants (N06A) (25.4%, n = 86), analgesics (N02) (20.7%, n = 70), and drugs used to treat diabetes (A10) (19.8%, n = 67) (Appendix A Figure A2).

3.2. LCA Model for Drug Family Prescription

The 20 indicators used for the LCA are described in detail in Appendix A (Table A1). Table 1 shows the fit criteria for each model. We observed that BIC indicated three classes, while aBIC increased this to four classes, with negligible differences with respect to the five classes (decrease of 0.826). Entropy was >0.8 in models with three or more classes. The pLMR values indicated four classes when considering the improvement of the other criteria as a whole. The adjustment for the pBRLT values was discarded because, as expected, their decline suggested more classes that would not contribute to clarification [29]. Moreover, the five-class model led to a small size profile (about 5% of the sample with 20 cases). Finally, the four-class model was chosen, allowing for a meaningful interpretation (Table 2).
According to the conditional probability shown by the indicators, we defined the class membership as follows: Class 1—patients with a small probability of having a drug prescription (labelled as low profile) (35.7% of patients, n = 122); Class 2—patients who had a high probability of having drug prescriptions for comorbidities from different domains (labelled as mixed profile) (28.7%, n = 98); Class 3—individuals who had a high probability of having prescriptions for cardio-metabolic-related medication (labelled as cardio-metabolic profile) (19.3%, n = 66); Class 4—patients with a high likelihood of using drugs for common psychiatric conditions (labelled as psychotropic profile) (16.4%, n = 56).
As a guide, the conditioned probabilities of prescribing therapeutic groups ≥0.5 are highlighted in bold in Table 2. In the low profile, a small conditional probability was found for all drugs, with the highest not exceeding 18% (antihypertensive drugs). For the mixed profile, there was no sharp break in probabilities. In this profile, it should be noted that the high probability of using analgesics (52%), followed by anxiolytic–hypnotic–sedatives (49%), antidepressants (42%), and anti-inflammatory drugs (37%), could be due more to osteo-articular rather than psychiatric problems. As cardiovascular treatments, only antihypertensive drugs were highly likely (66%). In the cardio-metabolic profile, a conditional probability ≥0.5 was observed for drugs prescribed for diabetes, antihypertensive drugs, lipid modifying agents, cardiac therapy drugs (0.45), and antithrombotic agents.

3.3. Inter-Class Comparison

Table 3 shows the inter-class comparisons of the patients’ demographic characteristics, family history of dementia, degree of medication, and use of high-scoring anticholinergic drugs. The sex co-variate exhibited significant differences: men were more frequently observed with a cardio-metabolic profile (53.0%), with a very high female incidence (82%) in those with a mixed profile (p = 0.000). Concerning the medication grade, there were significant differences (p = 0.000) in both the number of prescriptions and polypharmacy status. Thus, drug prescription was ≤5 in 96.7% (n = 118) of low profile members and in 62.5% (n = 35) of psychotropic profile members. Moderate polypharmacy (5–9 drugs) was present in the mixed profile (mean prescription 7.2 ± 2.9 drugs, CI95% 6.57–7.73) and in the cardio-metabolic profile (8.4 ± 3.5, CI95% 7.55–9.27), with 30.3% of class 3 members under excessive polypharmacy (≥10). The highest probability of taking prescription drugs for acid-related disorders was present in these two profiles. Significant differences (p = 0.021) were found in the number of drugs prescribed with anticholinergic action scored 3, the largest of which was observed in the psychotropic profile (16.1% of mem-bros).

3.4. Conversion

Of the sample, 33.3% of patients eventually developed dementia (n = 114), 84.1% (n = 96) of them were later re-diagnosed with AD (Alzheimer’s dementia), 2.63% (n = 3) with VaD (Vascular dementia), 2.63% (n = 3) with LBD (Lewy body dementia), and 10.53% (n = 12) with MVDA (mixed AD-VaD) (Table 4). The mean conversion interval was 2.55 ± 1.50 years (minimum 0.25, maximum 8.0). The cumulative annual frequency of progression was 21.1% in ≤1 year (n = 24); 50.9% in ≤2 years (n = 58), 75.4% in ≤3 years (n = 86); 87.7% in ≤4 years (n = 100); and 94.7% in ≤5 years (n = 108).
When comparing the converted patients vs. non-converted patients, significant differences were observed in terms of profile membership (p = 0.004 for all dementias, p = 0.003 for AD), age (p = 0.001, p = 0.005), and medication grade. Notably, for all dementias, 47.4% (n = 54) of the converted patients belonged to the low profile group vs. 29.8% (n = 68) of the non-converted; 12.3% (n = 14) of converted individuals were of the cardio-metabolic profile vs. 22.8% (n = 52) of the non-converted; 10.5% (n = 12) of re-diagnosed patients were aged <65 years vs. 27.6% (n = 63) of those who remained with MCI; 60.5% (n = 69) of the converted individuals were not under polypharmacy (<5 drugs) versus 45.2% (n = 103) of non-converted patients (p = 0.026). Similar data were found for AD (Table 4).
Finally, Table 5 records the results of the two binomial logistic regression analyses. The variables that did not give significant results are not shown. The low profile and <65 year groups were taken as references. The risk of progression to dementia was quadrupled in patients aged 65–79 years old (OR = 4.217, CI95% 2.092–8.500, p = 000) and tripled in patients aged >79 years old (OR = 2.945, CI95% 1.289–6.728, p = 0.010) compared to those aged <65 years. On the other hand, the mixed and cardio-metabolic profiles were protective compared to class 1 (OR = 0.421, CI95% 0.232–0.762, p = 0.004 and OR = 0.278, CI95% 0.137–0.565, p = 0.000, respectively). Regarding the conversion to AD, only the interval of 67–79 years was a risk factor for progression (OR = 3.259, CI95%1.614–6.583, p = 0.001), and all profiles were protective compared to class 1.

4. Discussion

In our study, we tried to clarify the drug prescription pattern at baseline in 342 patients diagnosed with MCI. In agreement with the literature, 33.3% progressed to AD within 5 years of onset (94.7%) [17,34]. Only 4.4% of these individuals (5.2–6.2%) developed dementia with a vascular component, also in line with the literature [17].
The four-class model from the LCA showed significant inter-cluster differences, indicating dominant morbidity patterns [35], which were summarized as musculoskeletal disorders (class 2), cardiovascular and metabolic diseases (class 3), and mental health disorders (class 4). Women were prevalent in the mixed profile (82.7% membership) and men were prevalent in the cardio-metabolic profile (53%) (p = 0.000), in line with studies in similar populations [36,37]. Regarding polypharmacy, a higher prevalence was observed compared to a study based on a Spanish primary care database: polypharmacy (≥5) was observed in 49.7% of our sample vs. 23.4% of the 65–79-year-old group and 36.7% of those aged ≥80 years; excessive polypharmacy (≥10) was observed in 11.1% of the sampled individuals vs. 2.61% of the 65–79-year-old group and 4.78% of those aged ≥80 years [38]. Thus, we cannot rule out that polypharmacy may have played a role in the cognitive frailty of our patients, 78.1% of whom were aged ≥65 years (mean 72.1 ± 9.5 years) (approximately 15–20% of people ≥65 years could have MCI) [39]).
Surprisingly, 47.4% of patients converted to dementia vs. 29.8% who did not convert (p = 0.004) (50.0% vs. 30.1% for AD, p = 0.003), which belonged to the low prescription profile. Additionally, the binomial regression analysis showed the other profiles to be protective against conversion. Moreover, patients with mixed and cardio-metabolic profiles were significantly (p = 0.000) subjected to polypharmacy (7.15 ± 2.89 and 8.41 ± 3.49 drugs prescribed, respectively) vs. the low profile group (1.35 ± 1.63). This was not a question of age, as it should be noted that the age difference between the profiles was at the limit of statistical significance (p = 0.05). It is true that, in the binomial regression model, age was the other risk factor for progression, which is to be expected: with <65 years as the reference group, there was a four-fold increase in risk at 65–79 years (for all dementias or AD) and a three-fold increase in risk at >80 years (for overall dementia).
Could the combination of pharmacological treatments targeting the underlying pathologies that are considered modifiable risk factors in middle and old age [20], which are also mutually reinforcing [40,41,42], was effective? In our “protective profiles”, the most frequently prescribed drugs were directed against hypertension, altered blood lipids, vascular protection, diabetes, pain (with its implications for physical inactivity and psychological disturbance) [43], depression, anxiety, and insomnia [44]. Most studies point to the association of polypharmacy or inappropriate medication with cognitive decline [11,45,46,47], even with daily consumption of ≥3 drugs [13]. The worsening of cognitive function attributable to antidepressants or benzodiazepines [48] argues against our finding of a protective role for the psychotropic profile. However, other authors reported that pharmacological intervention, even targeting cardiovascular risk factors, has no effect on the cognitive status [49]. For a similar home-dwelling population (74.7 ± 3.9 years, 63.9% females), no differences were found in medication grade between remaining or converted MCI, with the mean intake being comparable in both cases (4.08 ± 2.52 and 4.8 ± 3.8, respectively) to our results (4.31 ± 3.03) [50].
Nevertheless, a comprehensive study on the role of “protectan” combinations (vitamins, minerals, herbals) in preventing/delaying dementia, as opposed to complex drug associations, found that polypharmacy putatively facilitated cognitive decline while potentially favoring exposure to "protectans" [51]. They even found that the reversal of cognitive decline attributable to the Metabolic Enhancement for Neurodegeneration protocol (MEND) could be due to cardio-metabolic and anti-inflammatory medications, combined with the use of supplements and a healthy lifestyle. The methodological heterogeneity used to establish medication–cognitive impairment associations probably makes it difficult to contrast polypharmacy studies, contributing to the controversy.
As for the underlying comorbidities, one study noted that progression to AD was not influenced by the above-mentioned risk factors [52]. Therefore, let us assume that the converted patients could be in an early stage of AD before MCI diagnosis, and hence, the low profile membership (low comorbidity). However, it has been shown that about 80% of autopsied AD brains show compatible changes with other causes of dementia [18] and that, after decades-long latency periods from the initial pathological changes, the onset of symptoms is possibly triggered by environmental factors [53,54]. In this matter, recent LCA studies have shown that low morbidity profiles (7.5% diabetes, 8% depression, 25% hypertension, or 7% physical inactivity) exhibited a higher rate of MCI reversal [55].
Lastly, it has been emphasized that high standards of usual care [56] and appropriate treatment (especially of hypertension and diabetes) can prevent/delay the progression of MCI to dementia. Patients “hidden in the community”, without family and medical support, have a higher severity and prevalence of MCI [57]. Given the characteristics of the Spanish public healthcare system, most of our patients were referred to the neurologist from primary care services. The general population has regular access to these services and they are followed-up within prevention programs, although this entails polypharmacy. Perhaps our results reflect not so much the successful effect of prior medication, but of care, on the potential progression of MCI. Low drug prescription levels at baseline may reflect that these individuals have not seen their GP, by carelessness or because they did not need it. When they finally did, it was the symptoms were severe, because they, or their family, already perceived the cognitive impairment as suspicious of something more serious.

Limitations

Our study had some limitations, including the small sample size, the unavailability of data on the patients’ adherence to treatments, the lack of the starting date of each treatment, the unavailability of reversed cases, not including how long patients had been in MCI conditions prior to receiving the most accurate neurological diagnosis, and the unavailability of lifestyle parameters in the medical records. As we used written sources and not patient interviews, comorbidities were defined when patients had an altered parameter or were on treatment (e.g., hypertension, dyslipidemia, or diabetes). Thus, the same variable could have been repeated twice. Therefore, although they were collected (unpublished data), they were not included in this study.

5. Conclusions

At baseline, prior to the diagnosis of MCI, four drug prescription profiles were detected, indicating dominant morbidity patterns. In these patients who progressed to dementia according to the rates, intervals, and types reported in the literature, the profiles with a higher conditional probability of prescription (mixed, cardio-metabolic, and psychotropic) were shown to be protective against conversion to dementia, compared to the low profile. Their members were not significantly older, although age was certainly a risk factor. Beyond drug consumption, they could somewhat reflect previous patient care.
The classical three-step statistical analysis could be a tool for finding subjacent profiles in patients subjected to complex associations of variables. Based on baseline data collected from electronic medical records, this profiling could be explored at baseline as a predictor of the risk of dementia progression in MCI patients at onset.

Author Contributions

Conceptualization: C.G.-G. and M.Á.M.-M.; methodology: C.G.-G. and J.L.G.-C.; validation: C.G.-G., J.L.G.-C. and M.Á.M.-M.; formal analysis: J.L.G.-C. and M.J.T.-A.; investigation: M.J.T.-A. and J.F.-A.; resources: M.Á.M.-M.; data curation: J.L.G.-C.; writing—original draft preparation: M.J.T.-A. and C.G.-G.; writing—review and editing: C.G.-G.; funding acquisition: C.G.-G. All authors have read and agreed to the published version of the manuscript.

Funding

The publication of this study has been partially supported by the Research Group CTS-194 Bio-chemical Bases of Geriatrics and Gerontology, with funds of the University of Cadiz and the PAIDI (The Andalusian Research, Development and Innovation Program).

Institutional Review Board Statement

The study (MJT-MED-2019-01) was approved by the Comité de Ética de la Investigación de Cádiz (CEIC) ([email protected]) (HUPM, Av. Ana de Viya 21, 11009 CADIZ (SPAIN) (phone +0034956002100). The AEMPS (Agencia Española de Medicamentos y Productos Sanitarios, aemps.gob.es) qualified this study as EPA-OD, which does not require their approval. All the experiments were carried out in accordance with the relevant guidelines and regulations.

Informed Consent Statement

Routine data were extracted from anonymized medical records of long-term, non-interventional patients. Therefore, informed consent was waived.

Data Availability Statement

Data are not publicly available. Correspondence and requests for materials should be addressed to C.G.-G.

Acknowledgments

The online patient registers, resulting from neurological clinical practice supported by the Public Health System, are elaborated by the Department of Neurology (Hospital Universitario Puerta del Mar), associated with the University of Cadiz. The University of Cadiz provides their researchers with the software for statistical analysis.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Healthcare 11 02219 g0a1
Figure A1. Flowchart indicating procedure followed with the patients.
Figure A1. Flowchart indicating procedure followed with the patients.
Healthcare 11 02219 g0a1
Healthcare 11 02219 g0a2
Figure A2. Prescription frequency of drug families at baseline. § The group “antihypertensive drugs” includes C02AC—imidazoline receptor agonists, C03—diuretics, C07—beta-blocking agents, C08—calcium channel blockers, and C089—agents acting on the renin–angiotensin system. ¥ Non-steroid products. ‡ ADHD: attention deficit hyperactivity disorder.
Figure A2. Prescription frequency of drug families at baseline. § The group “antihypertensive drugs” includes C02AC—imidazoline receptor agonists, C03—diuretics, C07—beta-blocking agents, C08—calcium channel blockers, and C089—agents acting on the renin–angiotensin system. ¥ Non-steroid products. ‡ ADHD: attention deficit hyperactivity disorder.
Healthcare 11 02219 g0a2
Table A1. Description of indicators used in LCA that resulted from drug family regrouping.
Table A1. Description of indicators used in LCA that resulted from drug family regrouping.
1. Drugs used to treat diabetes (A10) §
  blood-glucose-lowering drugs (A10B)insulins and analogues (A10A)
2. Lipid-modifying agents (C10A)
  HMG CoA reductase inhibitors (C10AA)
  fibrates (C10AB)		other lipid-modifying agents (C10AX09)
3. Drugs used as antihypertensives
  Antiadrenergic agent (C02AC)
  diuretics (C03)
  beta blocking agents (C07)		calcium channel blockers (C08)
agents acting on the renin-angiotensin system (C09)
4. Cardiac therapy (C01)
  cardiac glycosides (C01A)
  antiarrhythmic, class I and III (C01B)		vasodilators (C01D), alone or combined
other cardiac preparations (C01E)
5. Antithrombotic agents (B01A)
  vitamin K antagonist (B01AA)
  heparin group (B01AB)		direct thrombin inhibitor (B01AE)
platelet aggregation inhibitors (B01AC)
6. Vasoprotectives (C04)/peripheral vasodilators (C05)
7. Antidepressants (N06A)
  SSRI † (N06AB)
  other mechanism (N06AX)		non-selective monoamine reuptake inhibitors (N06AA)
8. Anxiolytic–hypnotic–sedatives
  benzodiazepines: anxiolytic (N05BA); hypnotic–sedatives (N05CD)
  other hypnotic–sedatives, no barbiturates (N05CF, N05CM)
9. Antipsychotics (N05A)
  from all groups (N05AA to N05AX)
10. Drugs for other neurological uses
  anti-Parkinson drugs (N04)
  anti-epileptics (N03)		antivertigo preparations (N07C)
psychostimulants, drugs to treat ADHD, nootropics (N06B)
11. Drugs for paint treatment
  analgesic opioid (N02A) and other analgesic–antipyretics (N02B)
  other drugs used (N03, A03, DO4AB)
12. Anti-inflammatory products
  anti-inflammatory/antirheumatic products, non-steroidal (M01A) (M02A)
  glucocorticoids for systemic use (H02AB)
13. Osteoporosis (bisphosphonates [M05BA])
14. Drugs for acid related disorders (A02)
  antacids (A02A)proton pump inhibitors, H2-receptor antagonists (A02B)
15. Drugs for other digestive disorders
  drugs for functional gastrointestinal disorders (A03)
  drugs for constipation (A06)  
16. Drugs for urinary tract
  benign prostatic hypertrophy (G04C)       urinary frequency/incontinence (G04BD)  
17. Drugs for COPD-asthma
  obstructive airway diseases (R03)
  cough and cold preparation (R05) 		antihistamines for systemic use (R06)
18. Ophthalmologicals (S01)
19. Drugs for deficiencies
  vitamins and minerals (A11, 12)antianemic preparations (B12, folic acid, iron) (B03)
20. Thyroid therapy (H03)
§ in parentheses is the ACT nomenclature of the drug subgroup including in each item; † SSRI: selective serotonin reuptake inhibitor.

References

  1. Theou, O.; Rockwood, M.R.H.; Mitnitski, A.; Rockwood, K. Disability and co-morbidity in relation to frailty: How much do they overlap? Arch. Gerontol. Geriatr. 2012, 55, e1–e8. [Google Scholar] [CrossRef]
  2. Stuhec, M.; Gorenc, K.; Zelko, E. Evaluation of a collaborative care approach between general practitioners and clinical pharmacists in primary care community settings in elderly patients on polypharmacy in Slovenia: A cohort retrospective study reveals positive evidence for implementation. BMC Health Serv. Res. 2019, 19, 118. [Google Scholar] [CrossRef] [Green Version]
  3. Gutiérrez-Valencia, M.; Izquierdo, M.; Cesari, M.; Casas-Herrero, A.; Inzitari, M.; Martinez-Velilla, N. The relationship between frailty and polypharmacy in older people: A systematic review. Br. J. Clin. Pharmacol. 2018, 84, 1432–1444. [Google Scholar] [CrossRef] [Green Version]
  4. Jyrkkä, J.; Enlund, H.; Lavikainen, P.; Sulkava, R.; Hartikainen, S. Association of polypharmacy with nutritional status.; functional ability and cognitive capacity over a three-year period in an elderly population. Pharmacoepidemiol. Drug Saf. 2011, 20, 514–522. [Google Scholar] [CrossRef]
  5. Lai, S.W.; Lin, C.H.; Liao, K.F.; Su, L.T.; Sung, F.C.; Lin, C.C. Association between polypharmacy and dementia in older people: A population-based case-control study in Taiwan. Geriatr. Gerontol. Int. 2012, 12, 491–498. [Google Scholar] [CrossRef]
  6. Park, H.Y.; Park, J.W.; Song, H.J.; Sohn, H.S.; Kwon, J.W. The Association between Polypharmacy and Dementia: A Nested Case-Control Study Based on a 12-Year Longitudinal Cohort Database in South Korea. PLoS ONE 2017, 12, e0169463. [Google Scholar] [CrossRef] [Green Version]
  7. Moore, E.M.; Mander, A.G.; Ames, D.; Kotowicz, M.A.; Carne, R.P.; Brodaty, H.; Woodward, M.; Boundy, K.; Ellis, K.A.; Bush, A.I.; et al. Increased risk of cognitive impairment in patients with diabetes is associated with metformin. Diabetes Care 2013, 36, 2981–2987. [Google Scholar] [CrossRef] [Green Version]
  8. Wium-Andersen, I.K.; Osler, M.; Jørgensen, M.B.; Rungby, J.; Wium-Andersen, M.K. Antidiabetic medication and risk of dementia in patients with type 2 diabetes: A nested case-control study. Eur. J. Endocrinol. 2019, 181, 499–507. [Google Scholar] [CrossRef]
  9. Larsson, S.; Markus, H. Does Treating Vascular Risk Factors Prevent Dementia and Alzheimer’s Disease? A Systematic Review and Meta-Analysis. J. Alzheimer’s Dis. 2018, 64, 657–668. [Google Scholar] [CrossRef] [Green Version]
  10. Kodesh, A.; Sandin, S.; Reichenberg, A.; Rotstein, A.; Pedersen, N.L.; Ericsson, M.; Karlsson, I.K.; Davidson, M.; Levine, S.Z. Exposure to Antidepressant Medication and the Risk of Incident Dementia. Am. J. Geriatr. Psychiatry 2019, 27, 1177–1188. [Google Scholar] [CrossRef]
  11. Tapiainen, V.; Taipale, H.; Tanskanen, A.; Tiihonen, J.; Hartikainen, S.; Tolppanen, A.M. The risk of Alzheimer’s disease associated with benzodiazepines and related drugs: A nested case–control study. Acta Psychiatr. Scand. 2018, 138, 91–100. [Google Scholar] [CrossRef] [Green Version]
  12. Kristensen, R.U.; Nørgaard, A.; Jensen-Dahm, C.; Gasse, C.; Wimberley, T.; Waldemar, G. Polypharmacy and Potentially Inappropriate Medication in People with Dementia: A Nationwide Study. J. Alzheimer’s Dis. 2018, 63, 383–394. [Google Scholar] [CrossRef]
  13. Trevisan, C.; Limongi, F.; Siviero, P.; Noale, M.; Cignarella, A.; Manzato, E.; Sergi, G.; Maggi, S. Mild polypharmacy and MCI progression in older adults: The mediation effect of drug–drug interactions. Aging Clin. Exp. Res. 2021, 33, 49–56. [Google Scholar] [CrossRef]
  14. Rudolph, J.L.; Salow, M.J.; Angelini, M.C.; McGlinchey, R.E. The Anticholinergic Risk Scale and Anticholinergic Adverse Effects in Older Persons. Arch. Intern. Med. 2008, 168, 508–513. [Google Scholar] [CrossRef] [Green Version]
  15. Gray, S.L.; Anderson, M.L.; Dublin, S.; Hanlon, J.T.; Hubbard, R.; Walker, R.L. Cumulative Use of Strong Anticholinergic Medications and Incident Dementia. JAMA Intern. Med. 2015, 175, 401–407. [Google Scholar] [CrossRef] [Green Version]
  16. Tangalos, E.G.; Petersen, R.C. Mild Cognitive Impairment in Geriatrics. Clin. Geriatr. Med. 2018, 34, 563–589. [Google Scholar] [CrossRef]
  17. Mitchell, A.J.; Shiri-Feshki, M. Rate of progression of mild cognitive impairment to dementia: Meta-analysis of 41 robust inception cohort studies. Acta Psychiatr. Scand. 2009, 119, 252–265. [Google Scholar] [CrossRef]
  18. Schelke, M.W.; Attia, P.; Palenchar, D.J.; Kaplan, B.; Mureb, M.; Ganzer, C.A.; Scheyer, O.; Rahman, A.; Kachko, R.; Krikorian, R.; et al. Mechanisms of risk reduction in the clinical practice of Alzheimer’s Disease prevention. Front. Aging Neurosci. 2018, 10, 96. [Google Scholar] [CrossRef] [Green Version]
  19. Li, J.Q.; Tan, L.; Wang, H.F.; Tan, M.S.; Tan, L.; Xu, W.; Zhao, Q.F.; Wang, J.; Jiang, T.; Yu, J.T. Risk factors for predicting progression from mild cognitive impairment to Alzheimer’s disease: A systematic review and meta-analysis of cohort studies. J. Neurol. Neurosurg. Psychiatry 2016, 87, 476–484. [Google Scholar] [CrossRef] [Green Version]
  20. Livingston, G.; Huntley, J.; Sommerlad, A.; Ames, D.; Ballard, C.; Banerjee, S.; Brayne, C.; Burns, A.; Cohen-Mansfield, J.; Cooper, C.; et al. Dementia prevention.; intervention.; and care: 2020 report of the Lancet Commission. Lancet 2020, 396, 413–446. [Google Scholar]
  21. McKhann, G.; Drachman, D.; Folstein, M.; Katzman, R.; Price, D.; Stadlan, E.M. Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA work group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s disease. Neurology 1984, 34, 939–944. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  22. Román, G.C.; Tatemichi, T.K.; Erkinjuntti, T.; Cummings, J.L.; Masdeu, J.C.; Garcia, J.H.; Amaducci, L.; Orgogozo, J.M.; Brun, A.; Hofman, A.; et al. Vascular dementia: Diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993, 4, 250–260. [Google Scholar] [CrossRef]
  23. McKeith, I.G.; Dickson, D.W.; Lowe, J.; Emre, M.; O’brien, J.T.; Feldman, H.; Cummings, J.; Duda, J.E.; Lippa, C.; Perry, E.K.; et al. Diagnosis and management of dementia with Lewy bodies: Third report of the DLB Consortium. Neurology 2005, 65, 1863–1872. [Google Scholar] [CrossRef] [Green Version]
  24. Custodio, N.; Montesinos, R.; Lira, D.; Herrera-Pérez, E.; Bardales, Y.; Valeriano-Lorenzo, L. Mixed dementia: A review of the evidence. Dement. Neuropsychol. 2017, 11, 364–370. [Google Scholar] [CrossRef] [Green Version]
  25. Petersen, R.C. Mild cognitive impairment as a diagnostic entity. J. Intern. Med. 2004, 256, 183–194. [Google Scholar] [CrossRef]
  26. Gnjidic, D.; Hilmer, S.N.; Blyth, F.M.; Naganathan, V.; Waite, L.; Seibel, M.J.; McLachlan, A.J.; Cumming, R.G.; Handelsman, D.J.; Le Couteur, D.G. Polypharmacy cutoff and outcomes: Five or more medicines were used to identify community-dwelling older men at risk of different adverse outcomes. J. Clin. Epidemiol. 2012, 65, 989–995. [Google Scholar] [CrossRef]
  27. Durán, C.E.; Azermai, M.; Vander Stichel, R.H. Systematic review of anticholinergic risk scales in older adults. Eur. J. Clin. Pharmacol. 2013, 69, 1485–1496. [Google Scholar] [CrossRef]
  28. Groll, D.L.; To, T.; Bombardier, C.; Wright, J.G. The development of a comorbidity index with physical function as the outcome. J. Clin. Epidemiol. 2005, 58, 595–602. [Google Scholar] [CrossRef]
  29. Nylund, K.L.; Asparouhov, T.; Muthén, B.O. Deciding on the Number of Classes in Latent Class Analysis and Growth Mixture Modeling: A Monte Carlo Simulation Study. Struct. Equ. Model. 2007, 14, 553–569. [Google Scholar] [CrossRef]
  30. Sclove, S.L. Application of model-selection criteria to some problems in multivariate analysis. Psychometrika 1987, 52, 333–343. [Google Scholar] [CrossRef]
  31. Bray, B.C.; Lanza, S.T.; Collins, L.M. Modeling Relations Among Discrete Developmental Processes: A General Approach to Associative Latent Transition Analysis. Struct. Equ. Model. A Multidiscip. J. 2010, 17, 541–569. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  32. Morin, R.T.; Insel, P.; Nelson, C.; Butters, M.; Bickford, D.; Landau, S.; Saykin, A.; Weiner, M.; Mackin, R.S. Latent Classes of Cognitive Functioning among Depressed Older Adults without Dementia. J. Int. Neuropsychol. Soc. 2019, 25, 811–820. [Google Scholar] [CrossRef] [PubMed]
  33. Wang, J.; Wang, X. Mixture Modeling. In Structural Equation Modeling: Applications Using Mplus, 2nd ed.; Wang, J., Wang, X., Eds.; John Wiley: Hoboken, NJ, USA, 2019; pp. 339–442. [Google Scholar]
  34. Ward, A.; Tardiff, S.; Dye, C.; Arrighi, H.M. Rate of conversion from prodromal Alzheimer’s disease to Alzheimer’s dementia: A systematic review of the literature. Dement. Geriatr. Cogn. Dis. Extra 2013, 3, 320–332. [Google Scholar] [CrossRef] [PubMed]
  35. Larsen, F.B.; Pedersen, M.H.; Friis, K.; Glümer, C.; Lasgaard, M. A Latent Class Analysis of Multimorbidity and the Relationship to Socio-Demographic Factors and Health-Related Quality of Life. A National Population-Based Study of 162,283 Danish Adults. PLoS ONE 2017, 12, e0169426. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  36. Gutiérrez-Valencia, M.; Aldaz Herce, P.; Lacalle-Fabo, E.; Contreras Escámez, B.; Cedeno-Veloz, B.; Martínez-Velilla, N. Prevalence of polypharmacy and associated factors in older adults in Spain: Data from the National Health Survey 2017. Med. Clín. 2019, 153, 141–150. [Google Scholar] [CrossRef]
  37. Olaya, B.; Moneta, M.V.; Caballero, F.F.; Tyrovolas, S.; Bayes, I.; Ayuso-Mateos, J.L.; Haro, J.M. Latent class analysis of multimorbidity patterns and associated outcomes in Spanish older adults: A prospective cohort study. BMC Geriatr. 2017, 17, 186. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  38. Hernández-Rodríguez, M.; Sempere-Verdú, E.; Vicens, C.; González-Rubio, F.; Miguel-García, F.; Palop-Larrea, V.; Orueta-Sánchez, R.; Esteban-Jiménez, Ó.; Sempere-Manuel, M.; Arroyo-Aniés, M.P.; et al. Evolution of polypharmacy in a Spanish population (2005–2015): A database study. Pharmacoepidemiol. Drug Saf. 2020, 29, 433–443. [Google Scholar] [CrossRef]
  39. Alzheimer’s Association Report. 2017 Alzheimer’s disease facts and figures. Alzheimer’s Dement. 2017, 13, 325–373. [Google Scholar] [CrossRef]
  40. Katon, W.; Lyles, C.R.; Parker, M.M.; Karter, A.J.; Huang, E.S.; Whitmer, R.A. Association of depression with increased risk of dementia in patients with type 2 diabetes. The diabetes and aging study. Arch. Gen. Psychiatry 2012, 69, 410–417. [Google Scholar]
  41. Velly, A.M.; Mohit, S. Epidemiology of pain and relation to psychiatric disorders. Prog. Neuropsychopharmacol. Biol. Psychiatry 2018, 87, 159–167. [Google Scholar] [CrossRef]
  42. Verhagen, C.; Janssen, J.; Exalto, L.G.; van den Berg, E.; Johansen, O.E.; Biessels, G.J. Diabetes-specific dementia risk score (DSDRS) predicts cognitive performance in patients with type 2 diabetes at high cardio-renal risk. J. Diabetes Its Complicat. 2020, 34, 107674. [Google Scholar] [CrossRef]
  43. Soriano, J.B.; Rojas-Rueda, D.; Alonso, J.; Antó, J.M.; Cardona, P.-J.; Fernández, E.; Garcia-Basteiro, A.L.; Glenn, S.D.; Krish, V.; Lazarus, J.V.; et al. La carga de enfermedad en España: Resultados del Estudio de la Carga Global de las Enfermedades 2016. Med. Clin. 2018, 151, 171–190. [Google Scholar] [CrossRef]
  44. Forrester, N.S.; Gallo, J.J.; Smith, G.S.; Leoutsakos, J.M.S. Patterns of Neuropsychiatric Symptoms in MCI and Risk of Dementia. Am. J. Geriatr. Psychiatry 2016, 24, 117–125. [Google Scholar] [CrossRef] [Green Version]
  45. Risacher, S.L.; McDonald, B.C.; Tallman, E.F.; West, J.D.; Farlow, M.R.; Unverzagt, F.W.; Gao, S.; Boustani, M.; Crane, P.K.; Petersen, R.C.; et al. Association Between Anticholinergic Medication Use and Cogntition.; Brain Metabolism.; and Brain Atrophy in Cognitively Normal Older Adults. JAMA Neurol. 2016, 73, 721–732. [Google Scholar] [CrossRef] [Green Version]
  46. Kojima, T. The Need for Actions Against Polypharmacy in Older People with Frailty. Ann. Geriatr. Med. Res. 2018, 22, 111–116. [Google Scholar] [CrossRef] [Green Version]
  47. Heath, L.; Gray, S.L.; Boudreau, D.M.; Thummel, K.; Edwards, K.L.; Fullerton, S.M.; Crane, P.K.; Larson, E.B. Cumulative Antidepressant Use and Risk of Dementia in a Prospective Cohort Study. J. Am. Geriatr. Soc. 2018, 66, 1948–1955. [Google Scholar] [CrossRef] [PubMed]
  48. Ettcheto, M.; Olloquequi, J.; Sánchez-López, E.; Busquets, O.; Cano, A.; Manzine, P.R.; Beas-Zarate, C.; Castro-Torres, R.D.; García, M.L.; Bulló, M.; et al. Benzodiazepines and Related Drugs as a Risk Factor in Alzheimer’s Disease Dementia. Front. Aging Neurosci. 2020, 11, 344. [Google Scholar] [CrossRef] [PubMed]
  49. Ligthart, S.A.; Moll van Charante, E.P.; Van Gool, W.A.; Richard, E. Treatment of cardiovascular risk factors to prevent cognitive decline and dementia: A systematic review. Vasc. Health Risk Manag. 2010, 7, 775–785. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  50. Del Ser, T.; Zea, M.A.; Valentí, M.; Olazarán, J.; López-Álvarez, J.; Rebollo-Vázquez, A.; Ávila-Villanueva, M.; Frades, B.; Medina, M.; Fernández-Blázquez, M.A. Effects of commonly prescribed drugs on cognition and mild cognitive impairment in healthy elderly people. J. Psychopharmacol. 2019, 33, 965–974. [Google Scholar] [CrossRef] [PubMed]
  51. Mayburd, A.L.; Koivogui, M.; Baranova, A. Pharmacological signatures of the reduced incidence and the progression of cognitive decline in ageing populations suggest the protective role of beneficial polypharmacy. PLoS ONE 2019, 6, e0224315. [Google Scholar] [CrossRef] [Green Version]
  52. Eldholm, R.S.; Persson, K.; Barca, M.L.; Knapskog, A.-B.; Cavallin, L.; Engedal, K.; Selbaek, G.; Skovlund, E.; Saltvedt, I. Association between vascular comorbidity and progression of Alzheimer’s disease: A two-year observational study in Norwegian memory clinics. BMC Geriatr. 2018, 18, 120. [Google Scholar] [CrossRef] [PubMed]
  53. Alzheimer’s Association Report. 2021 Alzheimer’s disease facts and figures. Alzheimer’s Dement. 2021, 17, 327–406. [Google Scholar] [CrossRef]
  54. Sperling, R.A.; Aisen, P.S.; Beckett, L.A.; Bennett, D.A.; Craft, S.; Fagan, A.M.; Iwatsubo, T.; Jack Jr, C.R.; Kaye, J.; Montine, T.J.; et al. Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s Dement. 2011, 7, 280–292. [Google Scholar] [CrossRef] [Green Version]
  55. Katayama, O.; Lee, S.; Bae, S.; Makino, K.; Shinkai, Y.; Chiba, I.; Harada, K.; Shimada, H. Modifiable Risk Factor Possession Patterns of Dementia in Elderly with MCI: A 4-Year Repeated Measures Study. J. Clin. Med. 2020, 9, 1076. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  56. van Charante, E.P.M.; Richard, E.; Eurelings, L.S.; van Dalen, J.W.; Ligthart, S.A.; Van Bussel, E.F.; Hoevenaar-Blom, M.P.; Vermeulen, M.; van Gool, W.A. Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): A cluster-randomised controlled trial. Lancet 2016, 388, 797–805. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  57. Meguro, K.; Dodge, H.H. Vascular Mild Cognitive Impairment: Identifying Disease in Community-Dwelling Older Adults.; Reducing Risk Factors.; and Providing Support. The Osaki-Tajiri and Kurihara Projects. J. Alzheimer’s Dis. 2019, 70, S293–S302. [Google Scholar] [CrossRef]
Table 1. Fit parameters in latent class analysis.
Table 1. Fit parameters in latent class analysis.
ModelNparametersLL (df)AICBICaBICEntropypVLMRpBLRT
1 cl20−2964.3565968.7116045.4075981.963nanana
2 cl41−2782.8645647.7285804.9555674.8940.78700
3 cl62−2712.0925548.1845785.9425589.2650.8260.41180
4 cl83−2668.7765503.5525821.8425558.5470.8140.0270
5 cl104−2640.4065488.8125887.6325557.7210.8820.26530.004
6 cl125−2616.675483.7415963.0925566.5640.851 0.144
Table 2. Conditional probability of prescription according to drug family.
Table 2. Conditional probability of prescription according to drug family.
No. Individuals
(%)		Class 1
(Low Profile)		Class 2
(Mixed Profile)		Class 3
(Cardio-Metabolic Profile)		Class 4
(Psychological Profile)
Class prevalence (%) 35.728.719.316.4
No. individuals 122986656
Indicator Conditioned probability
Drugs used to treat diabetes75 (21.9)0.1290.1830.5250.113
Drugs used as antihypertensives167 (48.8)0.1810.6590.9180.304
Lipid-modifying agents102 (29.8)0.0940.2290.7670.291
Cardiac therapy43 (12.6)00.0780.4450.095
Antithrombotic agents101 (29.5)0.0950.1620.9280.205
Vasoprotectives/peripheral vasodilators28 (8.2)0.0270.0990.1240.113
Antidepressants88 (25.7)0.0370.4180.0880.606
Anxiolytic–hypnotic–sedatives122 (35.7)00.490.2860.913
Antipsychotics 21 (6.1)0.0150.0590.0360.188
Drugs for other neurological diseases25 (7.3)0.0350.0390.090.191
Pain treatment79 (23.1)0.050.520.2370.071
Anti-inflammatory products51 (14.9)0.0330.3660.1180.032
Bisphosphonates22 (6.5)0.0160.1810.0240
Drugs for acid-related disorders143 (41.8)0.030.740.8180.16
Drugs for other digestive disorders30 (8.8)00.2280.0770.025
Drugs for urinary tract40 (11.7)0.0440.1020.2950.085
Drugs for COPD-asthma53 (15.5)0.0350.2930.2250.069
Ophthalmologicals31 (9.1)0.0810.1270.1080.026
Drugs for deficiencies46 (13.5)0.0390.2880.1320.055
Thyroid therapy16 (4.7)0.0210.0790.0290.062
Drug prescription probability ≥0.5 is highlighted in bold to facilitate interpretation.
Table 3. Comparison of individual characteristics between classes.
Table 3. Comparison of individual characteristics between classes.
TOTALClass 1Class 2Class 3Class 4p *
No. individuals342122986656
Sex (female), n (%)225 (66.5)78 (63.9) 81 (82.7) 31 (47.0) 35 (63.6) 0.000
Age interval, n (%)
  <65 years75 (21.9)30 (24.6) 21 (21.4) 6 (9.1) 18 (32.1) 0.05
  65–79 years200 (58.5)69 (56.6) 61 (62.2) 45 (68.2) 25 (44.6)
  >79 years67 (19.6)23 (18.8) 16 (16.3) 15 (22.7) 13 (23.2)
Family history of dementia, n (%)51 (14.9)18 (14.8) 12 (12.2) 11 (16.7) 10 (17.9) 0.775
Medication grade
  No. of drugs (means ± SD)4.8 ± 3.8 1.4 ± 1.67.2 ± 2.98.4 ± 3.54.2 ± 1.80.000
  CI954.43–5.251.06–1.646.57–7.737.55–9.273.69–4.67
  Polypharmacy status (n,%)
     Non-polypharmacy, <5172 (50.3)118 (96.7) 14 (14.3) 5 (7.6) 35 (62.5) 0.000
     Moderate polypharmacy, 5–9132 (38.6)4 (3.3) 66 (67.3) 41 (62.1) 21 (37.5)
     Excessive polypharmacy, ≥1038 (11.1)0 (0.0) 18 (18.4) 20 (30.3) 0 (0.0)
Anticholinergic score 3, n (%) ¥ 28 (8.2) 5 (4.1) 11 (11.2) 3 (4.5) 9 (16.1) 0.021
* χ2 test; ¥ The prescribed drugs with an anticholinergic score of 3 (http://www.acbcalc.com accessed on 30 September 2021) were hydroxyzine, olanzapine, amitriptyline, paroxetine, solifenacin, tolterodine, trihexiphenidyl, and clomipramine.
Table 4. Characteristics of patients progressing to dementia.
Table 4. Characteristics of patients progressing to dementia.
CONVNCONVp *CONV AD †NCONV ADp *
TOTAL, n (%)114 (33.3)228 (66.7) 96 (28.1) 246 (79.9)
Class, n% §
  1 (low profile) 54 (47.4) 68 (29.8) 0.004 48 (50.0) 74 (30.1) 0.003
  2 (mixed profile) 26 (22.8) 72 (31.6) 25 (26.0) 73 (29.7)
  3 (cardio-metabolic profile) 14 (12.3) 52 (22.8) 10 (10.4) 56 (22.8)
  4 (psychotropic profile) 20 (17.5) 36 (15.8) 13 (13.5) 43 (17.5)
Sex (female), n (%)80 (70.2)145 (63.6)0.24868 (70.8)157 (63.8) 0.242
Age interval, n (%)
  <6512 (10.5)63 (27.6)0.00112 (12.5)63 (25.7) 0.005
  65–7979 (69.3)121 (53.1) 69 (71.9)131 (53.5)
  >7923 (20.2)44 (19.3) 15 (15.6)52 (20.8)
Family history of dementia, n (%)15 (13.2)36 (15.8)0.52014 (14.6)37 (15.4)0.915
Medication grade
  No. of drugs (means ± SD)4.1 ± 3.75.2 ± 3.80.0144.0 ± 3.85.2 ± 3.80.014
   CI953.44–4.804.69–5.70 3.27–4.804.68–5.63
  Polypharmacy status, n (%)
    <5 drugs69 (60.5)103 (45.2)0.02659 (61.5)113 (45.9) 0.004
    5–9 drugs36 (31.6)96 (42.1) 29 (30.2)103 (41.9)
    ≥10 drugs9 (7.9)29 (12.7) 8 (8.3)30 (12.2)
Anticholinergic score 3, n (%) 10 (8.8) 18 (7.9) 0.7807 (7.3) 21 (8.5) 0.706
* χ2 test; § cluster membership is considered an independent variable; CONV: patients converted to dementia; NCONV: patients not converted to dementia; CONV AD: † patients converted to AD (without vascular component); NCONV AD: remaining individuals.
Table 5. Binomial regression logistic model to assess the risk of MCI conversion to dementia.
Table 5. Binomial regression logistic model to assess the risk of MCI conversion to dementia.
Coef. SEWaldgpORCI95%
ALL DEMENTIAS
<65 years old referent 1
65–79 years old1.4390.35816.19710.0004.2172.092–8.500
>79 years old1.0800.4216.57010.0102.9451.289–6.728
class 1 referent 15.99330.001
Class 2−0.8660.3038.17610.0040.4210.232–0.762
Class 3−1.2810.36212.53010.0000.2780.137–0.565
Class 4−0.2950.3510.70710.4000.7440.374–1.481
Constant−1.3040.34314.43710.0000.271
AD
<65 years old referent 1
65–79 years old1.1820.35910.85210.0013.2591.614–6.583
>79 years old0.5040.4461.28210.2581.6560.692–3.966
class 1 referent 16.27430.001
Class 2−0.7330.3055.77910.0160.4800.264–0.873
Class 3−1.4800.39913.75610.0000.2280.104–0.498
Class 4−0.7830.3844.15010.0420.4570.215–0.971
Constant−1.2160.34212.65710.0000.296
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Gómez-Gómez, C.; Moya-Molina, M.Á.; Tey-Aguilera, M.J.; Flores-Azofra, J.; González-Caballero, J.L. Baseline Profiles of Drug Prescriptions Prior to Diagnosis of Mild Cognitive Impairment (MCI) Obtained by Latent Class Analysis (LCA), and Assessment of Their Association with Conversion to Dementia. Healthcare 2023, 11, 2219. https://doi.org/10.3390/healthcare11152219

AMA Style

Gómez-Gómez C, Moya-Molina MÁ, Tey-Aguilera MJ, Flores-Azofra J, González-Caballero JL. Baseline Profiles of Drug Prescriptions Prior to Diagnosis of Mild Cognitive Impairment (MCI) Obtained by Latent Class Analysis (LCA), and Assessment of Their Association with Conversion to Dementia. Healthcare. 2023; 11(15):2219. https://doi.org/10.3390/healthcare11152219

Chicago/Turabian Style

Gómez-Gómez, Carmen, Miguel Ángel Moya-Molina, Manuel Jesús Tey-Aguilera, Jorge Flores-Azofra, and Juan Luis González-Caballero. 2023. "Baseline Profiles of Drug Prescriptions Prior to Diagnosis of Mild Cognitive Impairment (MCI) Obtained by Latent Class Analysis (LCA), and Assessment of Their Association with Conversion to Dementia" Healthcare 11, no. 15: 2219. https://doi.org/10.3390/healthcare11152219

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop