International Journal of Public Health ORIGINAL ARTICLE published: 18 October 2023 doi: 10.3389/ijph.2023.1606259 Shifts in Chronic Disease Patterns Among Spanish Older Adults With Multimorbidity Between 2006 and 2017 Jeroen J. A. Spijker * and Elisenda Rentería Centre d’Estudis Demogràfics, Bellaterra, Spain Objectives: To investigate changes in multimorbidity patterns among Spanish older adults. Methods: Data come from the Spanish National Health Survey (ENSE) for individuals aged 60–89 years (2006: n = 9,758; 2017: n = 8,535). Prevalence rates and relative risks of 20 chronic conditions are estimated for the multimorbidity (3+ chronic conditions) sample, along with observed-to-expected prevalence of three-way disease combinations. Principal component and cluster analyses identify multimorbidity patterns and track temporal changes. Results: Overall, multimorbidity remained stable [2006: 59.6% (95% CI: 58.7%–60.6%); 2017: 60.3% (CI: 59.3%–61.3%)], except at older ages. Women exhibited higher multimorbidity prevalence, but sex differences declined by five percentage points. Low-high education differences widened by three percentage points. In 2017 most individuals living with multimorbidity experienced hypertension (63.4%), osteoarthrosis (62.4%) and chronic back pain (55.9%). These chronic conditions also dominate the most common triadic combinations. Multimorbid men also saw increases in cholesterol and diabetes. Edited by: Saverio Stranges, Conclusion: Multimorbidity trends and the most common combination of diseases can Western University, Canada help plan healthcare for an ageing population. Sex and socioeconomic differences pose Reviewed by: additional public health challenges as women and deprived populations tend to have more Kathryn Nicholson, Western University, Canada health complexities. *Correspondence: Keywords: multimorbidities, morbidity, chronic disease, ageing, health inequalities, cluster analysis, principal Jeroen J. A. Spijker component analysis, Spain jspijker@ced.uab.es Received: 29 May 2023 INTRODUCTION Accepted: 28 September 2023 Published: 18 October 2023 Life expectancy in Spain has steadily increased in recent decades and now ranks among the world’s Citation: highest [1]. As deaths progressively occur at older ages, important changes in the type of diseases and Spijker JJA and Rentería E (2023) chronic conditions have ocurred. This development is best described by the epidemiological Shifts in Chronic Disease Patterns transition theory [2]. A key concern with increasing longevity is whether additional years are Among Spanish Older Adults With Multimorbidity Between spent in good health or if morbidity is becoming more prevalent. To date, in low-mortality countries, 2006 and 2017. no clear pattern has been established between life expectancy and disease- or disability-free life Int J Public Health 68:1606259. expectancy (DFLE) [3–11]. Results depend on several factors, including the chosen measure for doi: 10.3389/ijph.2023.1606259 morbidity when calculating DFLE (e.g., disability, chronic conditions) and which country and time Int J Public Health | Owned by SSPH+ | Published by Frontiers 1 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity TABLE 1 | List of chronic conditions used in the analyses according to prevalence rank order of the total sample in 2017 (%), the prevalence in the total sample, the multimorbidity sample and the relative risk for multimorbidity, 2006 and 2017 by gender for ages 60–89 (Spain. 2006 and 2017). Chronic Total Men Women Total Men Women condition 2006 2017 2006 2017 2006 2017 2006 2017 2006 2017 2006 2017 TS MMS TS MMS TS MMS TS MMS TS MMS TS MMS RR RR RR RR RR RR 1 Hypertension 45.0 57.9 47.7 63.4 39.7 54.4 47.6 66.3 49.3 60.0 47.8 61.5 2.2 2.7 2.2 2.5 2.3 3.0 2 Osteoarthrosis 49.2 69.5 43.1 62.4 33.7 54.8 27.9 44.6 61.6 78.0 55.9 74.4 3.6 4.6 4.2 5.0 2.9 3.9 3 Cholesterol 27.5 38.5 38.5 52.4 24.3 37.3 38.0 55.2 30.0 39.3 38.8 50.4 3.5 3.0 3.2 3.0 3.7 3.2 4 Chronic back 38.5 57.8 37.5 55.9 27.4 46.5 28.1 44.7 47.5 64.3 45.5 63.3 5.8 5.8 5.4 4.8 5.5 6.3 pain 5 Obesity 23.0 31.0 23.8 32.8 21.0 31.1 22.5 32.5 24.6 30.9 24.9 32.9 2.7 3.1 2.9 3.0 2.6 3.3 6 Diabetes 16.1 21.8 20.5 29.5 17.3 26.1 22.9 35.2 15.1 19.3 18.4 25.6 2.8 4.3 3.0 3.9 3.0 6.0 7 Varicose veins 20.8 30.9 17.4 26.2 10.3 17.6 8.7 13.9 29.2 38.6 24.7 34.5 5.3 6.8 5.6 4.8 4.2 6.8 8 Mental health 17.6 26.8 16.1 24.3 9.4 16.6 8.8 14.3 24.2 32.8 22.2 31.0 6.9 6.7 7.0 5.5 5.7 6.4 problems 9 Heart disease 13.7 20.3 14.9 21.6 14.6 24.3 16.5 26.0 12.9 18.0 13.5 18.6 5.2 4.6 4.7 4.6 8.3 5.4 10 Allergies 8.1 12.2 12.5 18.2 6.1 10.3 9.2 14.6 9.7 13.3 15.3 20.7 6.0 4.8 5.2 4.7 6.3 4.5 11 Urinary 10.3 16.0 11.5 18.2 9.3 16.9 9.8 17.0 11.0 15.5 13.0 18.9 9.2 12.6 9.1 10.2 9.8 16.2 incontinence 12 COPD/asthma 11.7 17.3 10.4 15.4 13.9 23.1 11.0 17.5 10.0 13.9 9.8 14.0 4.9 5.6 4.7 4.8 7.6 8.4 13 Osteoporosis 13.7 21.7 10.2 15.6 3.6 7.0 2.1 3.8 21.7 30.3 17.0 23.5 13.1 7.9 29.1 16.5 8.7 5.8 14 Prostate 9.3 13.1 8.6 12.2 21.0 35.6 18.8 30.3 — — — — 3.5 3.8 5.3 5.1 — — 15 Thyroid 4.6 6.8 8.1 11.8 1.1 1.9 2.5 4.0 7.5 9.7 12.9 17.0 5.1 4.7 6.3 5.6 3.7 3.6 dysfunction 16 Haemorrhoids 9.3 14.5 7.9 12.4 7.8 14.2 6.0 10.4 10.5 14.7 9.6 13.7 8.8 10.4 9.0 9.3 8.6 10.9 17 Migraine 10.7 16.8 7.5 11.3 6.1 11.0 4.1 6.8 14.3 20.2 10.3 14.3 11.0 6.6 7.9 6.0 11.9 6.0 18 Chronic skin 7.0 10.9 7.4 11.4 7.8 14.3 7.4 12.6 6.4 8.9 7.5 10.6 8.4 8.1 10.4 8.3 7.5 8.3 problems 19 Constipation 10.2 15.9 7.2 11.4 6.6 11.0 3.7 6.6 13.1 18.7 10.1 14.6 8.4 14.1 4.8 15.8 13.7 11.5 20 Gastric and 6.0 8.8 4.2 6.5 6.0 9.5 3.9 6.4 6.1 8.4 4.5 6.5 4.5 8.2 3.9 5.9 6.6 14.7 duodenal ulcer Notes: TS, total sample; MMS, multimorbid sample; RR, relative risk. RRs equal or greater than 10.0 are printed in bold. These chronic conditions have been diagnosed by a medical doctor and the respondent experienced during the last 12 months. The exception is obesity, as this was calculated on the basis of weight and height of the respondent, taking a BMI of at least 30 as the definition of obesity. Respondents are 60–89 years old. Regarding mental health problems [8], in 2006 these pertain to the chronic condition category “depression, anxiety or other mental illnesses.” In 2017 this was split up into separate categories (“depression,” “anxiety,” “other mental illnesses”). Hence, in order to establish a comparable list of health conditions for the analysis, the three distinct mental health categories from the 2017 data were combined into one, thus treating respondents who reported multiple mental health issues as having only one mental health problem. However, considering the congruence of these categories with the 2006 classification and the marginal shifts in proportions over time (see table), results are unlikely to be affected. Lastly, ENSE also asked respondents about several other chronic conditions, including stroke, cancer and menopause, but as they had a prevalence too we considered too low for our triadic combination analysis (<4%) we did not analyse them. Neither did we analyse cataracts despite a 16% prevalence in 2017 because of the likelihood of being cured if operated (meaning that it is no longer a chronic condition). Data source: the Spanish National Health Survey (ENSE). period is analysed. In Spain, despite declines in severe disability this context, despite Spain’s high life expectancy, adults aged among older people during the 1990s [10–12], self-care disability 70 years and older exhibited higher rates of multimorbidity among women underwent a sharp rise [10]. From 1997 to 2010, compared to other European countries in 2011 [18]. disease presence increased due to earlier age-specific onset and Multimorbidity studies often adopt varying approaches. Some increased life expectancy [13] and the most recent data suggest a exclusively focus on the coexistence of two diseases/chronic continuation of this trend [14]. conditions (henceforth comorbidity) or three or more diseases/ It appears that declines in mortality have increased the time chronic conditions (henceforth multimorbidity) in different spent in disability and with chronic diseases. However, when regions and from different surveys. These studies analyse age examining the epidemiological transition, the focus tends to be on and cohort patterns, socioeconomic inequalities or link single cause-of-death groups or the presence of a chronic multimorbidity to quality of life and healthcare use [19–25]. condition/disease (e.g., [14, 15] for Spain), rather than In contrast, other studies emphasize the nature of associations multiple causes of death or the simultaneous presence of among diseases, necessitating the application of distinct cluster or chronic conditions/diseases. factor analysis methods to discern the most prevalent disease Furthermore, as health inequalities continue to exist between combinations [25–29]. men and women and social classes [16], it becomes crucial to Regarding common co- and multimorbidity patterns, Garin explore, besides the prevalence of specific diseases across sex and et al [28] identified distinct patterns in the population aged socioeconomic groups (SES), the co-occurrence of chronic 50 years and older across nine countries, focussing on conditions. This is because multimorbidity is strongly 12 highly prevalent chronic conditions. Hypertension emerged associated with increased frailty, risk of disability and, as the most prevalent comorbid condition, especially among therefore, more healthcare utilisation and risk of death [17]. In individuals with obesity, stroke, diabetes, and angina. Arthritis Int J Public Health | Owned by SSPH+ | Published by Frontiers 2 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity and cataracts were also common comorbid conditions. In Spain, combinations. The observed prevalence was obtained from the analysis of data from the COURAGE project revealed the presence of survey, and the expected prevalence was estimated by multiplying a “cardio-respiratory” (angina, asthma, COPD) and “metabolic” the prevalence of each of the three single conditions. An O/E ratio (diabetes, obesity, hypertension) pattern. Another study using the greater than 1 indicates that diseases are not independent, as the same dataset [27], which also included anxiety disorders in the combination occurs more frequently than expected. Sample weights analysis, highlighted a clear “mental-articular” pattern (arthritis, were applied to all prevalence estimates to account for age, sex and depression, anxiety), possibly due to the psychological burden of regional population structure. The prevalence of multimorbidity and arthritis [30]. triadic disease combinations was computed for men, women, Concerning SES differentials in multimorbidity, studies from 10 years age groups (60–69, 70–79, 80–89) and educational other countries have identified a gradient similar to that of mortality, attainment (less than primary school, primary and lower with a higher prevalence of more than two diseases among lower secondary school, upper secondary, vocational training and socioeconomic groups [31–33]. However, a similar study in Spain is university). needed to better understand which diseases are more affected by SES. In this context, our paper’s main objective is to analyse trends in Statistical Analysis multimorbidity prevalence among older people, defined here as the To facilitate our aim of identifying meaningful groups of co- presence of at least three chronic conditions among 60–89 years-olds occurring chronic conditions, we exclusively analysed the living in Spain in 2006 and 2017. We also take special interest in multimorbidity sample (mm-sample). To enhance the assessing sex, age and socioeconomic health differences and whether reliability of our findings, we employed two methods: there have been changes in the patterns of co-occurring common chronic diseases over time. Similar to Islam et al. [34], we apply i) Principal component analysis (PCA) of the 20 chronic different methods to a single dataset as there is no consensus in conditions with a varimax rotation to simplify the existing literature on how to measure the co-occurrence of diseases. complex relationships among the different chronic However, if our results exhibit a high degree of consistency, it would conditions (variables) into a reduced set of factors provide confidence in the observed patterns. Specifically, we first (principal components). Each component is a weighted identify the most important combinations of three chronic sum of the original variables, where the weights (factor conditions (referred to as triadic combinations) and then employ loadings) indicate the correlations between the variable cluster analysis and principal component analysis on 20 chronic and the principal component. Factors were selected based conditions among the older population living with multimorbidity. on eigenvalues greater than one and variable factor loadings greater than 0.3. Components were ordered by their ability to explain the maximum possible variance. The component METHODS with the highest eigenvalue was considered the most important in explaining the observed data variation [37]. Study Design and Population ii) Agglomerative hierarchical cluster analysis, based on the Data on the Spanish population aged 60–89 years are extracted from relative proximity of the 20 chronic conditions. The the 2006 and 2017 Spanish National Health Surveys (ENSE), with conventional approach to cluster analysis classifies n sample sizes of 9,758 and 8,535, respectively. The ENSE is a cross- respondents into clusters based on proximity, producing an sectional survey conducted roughly every 5 years on a representative n × n proximity matrix reflecting the degree of closeness sample of the non-institutionalised population. In 2006, a among respondents. However, it is also possible to produce comparable list of chronic conditions was introduced, allowing multimorbidity groups (clusters) based on the presence or co- for comparisons with the most recent data available from the occurrence of two or more chronic conditions in a single same survey. Individuals were considered to have a chronic individual by reducing the data matrix to a much smaller p × p condition if they reported being diagnosed with and suffering proximity matrix of chronic conditions [34, 38]. Following the from the condition during the last 12 months. We define methodology applied by Islam et al [34], we used Yule’s Q as multimorbidity as the presence of at least three chronic the similarity measure. To measure distance between two conditions from the list of 20 conditions used in our analysis (see clusters, we opted for centroid linkage as it yielded a clustering Supplementary Table S1 for sample description and Table 1 for the solution similar to the components obtained from PCA. list of chronic conditions). This threshold has also been used in other Analyses were performed using SPSS version 24. studies [34–36]. Outcome Variables RESULTS For each chronic condition, prevalence rates were calculated for the multimorbidity sample (mm-sample) as well as the Relative Risk (RR). RR represents the likelihood of multimorbid persons Time Changes in Prevalence Rates in managing a specific chronic condition compared to persons Multimorbidity According to without multimorbidity but who live with the same condition Sociodemographics [35]. To investigate the independence of diseases, observed-to- Between 2006 and 2017, there was no overall change in the expected prevalence ratios (O/E) were calculated for all triadic prevalence of multimorbidity for people aged 60–89 years. In Int J Public Health | Owned by SSPH+ | Published by Frontiers 3 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity FIGURE 1 | Proportion of the Spanish population aged 60–89 years who live with multimorbidity by single age & sex, Spain 2006 vs. 2017. Note: The age-specific proportions were fitted in EXCEL using a logarithmic FIGURE 2 | Proportion of the Spanish population aged 60–89 who live equation. Data source: the Spanish National Health Survey (ENSE). with multimorbidity by sex & educational attainment, Spain 2006 vs. 2017. Note: The proportions are age-standardised to take into account the educational expansion that took place in Spain between the late 1950s and 1970s (i.e., correcting for the fact there are more older people who had both years, 6 out of every 10 people had at least three chronic not completed primary school in 2006 than in 2017, who are also more likely to conditions [2006: 59.6% (95% CI: 58.7%–60.6%); 2017: 60.3% live with 3+ conditions). This permits results to be compared over time. Data source: the Spanish National Health Survey (ENSE). (CI: 59.3%–61.3%)]. The prevalence of multimorbidity was notably lower for men, although the sex gap reduced by five percentage points over time. Prevalence among men increased by (Table 2). Results showed that among those living with at three percentage points to 53.1% (95% CI: 51.6%–54.7%), while least three chronic conditions, women managed on average among women it decreased by two percentage points to 66.3% 0.7 chronic conditions more than men did in 2006 [women (95% CI: 65.0%–67.7%). Regarding age, multimorbidity was 5.34 (95% CI: 5.27–5.40); men: 4.64 (95% CI: 4.60–4.67)] and higher in 2017 than in 2006 among men aged 64 years and 0.8 more in 2017 [women 5.41 (95% CI: 5.33–5.48); men: 4.60 older and women aged 70 years and older. However, below (95% CI: 4.57–4.64)]. For both sexes the higher the educational these ages, multimorbidity declined (Figure 1). Changes were level, the fewer chronic conditions. There were no statistically not uniform across educational levels and sex as increases over significant increases in the number of chronic conditions among time occurred mainly among lower-educated men, while among multimorbid individuals between the two periods, irrespective of higher-educated Spanish women there was a slight decline their educational category. For individuals without between periods (Figure 2). multimorbidity, there were no statistically significant We then estimated the average number of conditions among differences in the average number of conditions in the two individuals with and without multimorbidity by sex and survey years, regardless of sex and educational attainment. The educational level and compared their trend over time only exceptions were in 2006 [all women: 1.30 (95% CI: TABLE 2 | Average number of conditions together with 95% confidence intervals among persons who live without/with multimorbidity by sex and educational level (age- standardized) (Spain. 2006 and 2017). Men Women Multimorbidity Multimorbidity No Yes No Yes 2006 2017 2006 2017 2006 2017 2006 2017 Total 1.15 1.17 4.64 4.60 1.30 1.19 5.34 5.41 1.11–1.18 1.14–1.21 4.60–4.67 4.57–4.64 1.22–1.38 1.12–1.27 5.27–5.40 5.33–5.48 Less than primary school 1.12 1.21 4.94 5.05 1.38 1.32 5.67 5.83 1.04–1.19 1.12–1.31 4.89–5.00 4.97–5.13 1.21–1.54 1.14–1.49 5.57–5.77 5.69–5.97 Primary/lower sec school 1.20 1.20 4.59 4.52 1.29 1.19 5.21 5.26 1.15–1.25 1.15–1.25 4.55–4.64 4.47–4.57 1.19–1.40 1.08–1.30 5.12–5.29 5.16–5.36 Upper sec/vocational/university 1.11 1.15 4.36 4.46 1.20 1.10 4.76 5.03 1.03–1.18 1.08–1.21 4.27–4.45 4.38–4.53 1.02–1.39 0.96–1.25 4.56–4.96 4.86–5.20 “Total” written in bold. Data source: the Spanish National Health Survey (ENSE). Int J Public Health | Owned by SSPH+ | Published by Frontiers 4 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity 1.22–1.38); all men 1.15 (95% CI: 1.11–1.18)] and for those with samples. Despite similarity in ranking in both years, noteworthy less than primary school [1.12 (95% CI: 1.04–1.19) for men and is the increase in RR for multimorbidity between 2006 and 1.38 (95% CI: 1.21–1.54) for women]. Age differences were also 2017 for constipation, especially for men (from 4.8 to 15.8), analysed, but no or only minor increases over age were discerned. while marked increases among women were observed for urinary For instance, among men with less than primary school, incontinence (from 9.8 to 16.2) and duodenal ulcers (from 60–64 years-olds had an average of 1.25 chronic conditions 6.6 to 14.7). compared to 1.19 for 85–89 years-olds. The corresponding figures for low-educated women were 1.29 and 1.44, Prevalence of Triadic Combinations respectively. For those with upper secondary school and Our next step involved analysing the triadic combinations of higher education, the figures were 1.04 and 1.24 for 60–64 and chronic conditions among multimorbid individuals (Table 3). 85–89 years-old men and 0.93 and 1.37 chronic conditions for The 20 selected conditions provided us with 20!/(17!·3!) = 60–64 and 85–89 years-old women (see Supplementary 1,140 combinations. After sorting them according to frequency Figure S1). of occurrence in 2017 and considering only the 10 top-ranked combinations for the total population aged 60–89 years, we observe that triadic combinations involving osteoarthrosis [8], Prevalence of Chronic Diseases and the hypertension [6] and back pain [6] predominated in 2017. This Relative Risks of Multimorbidity pattern remained stable between 2006 and 2017 among women, Table 1 presents the prevalence of the chronic conditions for both however, among men, there was a clear shift towards more the total and multimorbidity samples. In 2017, the highest multimorbidity associated with hypertension (+4), cholesterol prevalences among the multimorbidity sample in 2017 were (+4) and diabetes (+2). Further analysis considering education observed for hypertension (63.4%), osteoarthrosis (62.4%), levels revealed that a higher proportion of older men living with chronic back pain (55.9%) and cholesterol (52.4%). Cholesterol multimorbidity experienced problems associated with the exhibited the largest increase since 2006 (+13.9 percentage prostate in 2006 compared to 2017. Regarding women, bone points/36% higher). Regarding sex differences and changes diseases, back pain, varicose veins and mental health issues over time, men showed the largest increases in cholesterol remained important conditions in the multimorbidity patterns (+17.9 percentage points/48% higher), followed by over time, especially for those with lower secondary school or less hypertension (+11.9 percentage points/22% higher) and (Supplementary Table S2). diabetes (+9.1 percentage points/35% higher). Among women, Table 3 also displays the estimated O/E ratio for the most remarkable increases were noted for cholesterol frequent triadic combinations of chronic conditions. Notably, (+11.1 percentage points/28% higher), allergies suffering from the three most frequent triadic combinations (+7.4 percentage points/56% higher) and thyroid dysfunction (hypertension/osteoarthrosis/back pain, hypertension/ (+7.3 percentage points/75% higher). Concomitantly, noteworthy cholesterol/osteoarthrosis, and cholesterol/osteoarthrosis/back declines were observed in osteoarthrosis among men pain) is approximately as expected based on the prevalence of (−7.1 percentage points/19% lower) and osteoporosis among the individual conditions (O/E ratio between 0.9 and 1.2). On women (−6.8 percentage points/22.4% lower). The table also the other hand, the O/E ratio of men suffering from shows the RR for specific chronic conditions by comparing the osteoarthrosis and backpain along with either mental health prevalence of each condition between the mm- and non-mm or varicose veins equaled, respectively 1.6 and 1.9 in 2017, TABLE 3 | Prevalence (prevalence rank) and observed/expected ratio of the most common 3-way disease combinations of chronic conditions among 60–89 years-old people living with multimorbidity in Spain, 2006 and 2017. Total and by sex. Combination Total Men Women Total Men Women 2006 2017 2006 2017 2006 2017 2006 2017 2006 2017 2006 2017 Prevalence (rank) Prevalence (rank) Prevalence (rank) O/E (Rank) O/E (Rank) O/E (Rank) Hypertension & osteoarthrosis & back pain 25.1 (1) 25.2 (1) 15.2 (1) 15.4 (3) 30.9 (1) 31.8 (1) 1.1 1.1 1.1 1.2 1.0 1.1 Hypertension & cholesterol & osteoarthrosis 16.0 (3) 21.0 (2) 11.0 (4) 15.5 (2) 18.9 (6) 24.7 (3) 1.0 1.0 1.0 0.9 1.0 1.1 Cholesterol & osteoarthrosis & back pain 17.6 (2) 20.3 (3) 11.7 (2) 13.0 (6) 21.1 (3) 25.2 (2) 1.1 1.1 1.2 1.2 1.1 1.1 Hypertension & cholesterol & back pain 12.2 (10) 18.1 (4) 8.1 (8) 15.2 (4) 14.7 (12) 20.1 (4) 0.9 1.0 0.8 0.9 1.0 1.0 Hypertension & cholesterol & diabetes 6.6 (43) 14.3 (5) 6.2 (21) 17.9 (1) 6.8 (64) 11.8 (19) 1.3 1.5 1.2 1.4 1.5 1.5 Osteoarthrosis & back pain & varicose veins 15.9 (4) 13.9 (6) 5.8 (26) 5.3 (38) 21.8 (2) 19.8 (5) 1.3 1.5 1.3 1.9 1.1 1.2 Obesity & osteoarthrosis & back pain 13.8 (7) 13.2 (7) 8.9 (7) 7.5 (18) 16.7 (8) 17.0 (7) 1.1 1.2 1.1 1.2 1.1 1.1 Osteoarthrosis & back pain & mental health 15.2 (5) 13.0 (8) 6.3 (20) 4.4 (55) 20.3 (5) 18.9 (6) 1.4 1.5 1.5 1.6 1.2 1.3 Hypertension & Obesity & osteoarthrosis 13.7 (8) 13.0 (9) 9.9 (5) 8.8 (11) 15.9 (9) 15.9 (10) 1.1 1.0 1.1 0.9 1.1 1.1 Hypertension & diabetes & osteoarthrosis 9.3 (17) 12.2 (10) 6.7 (15) 9.8 (9) 10.7 (22) 13.8 (12) 1.1 1.0 0.9 0.9 1.2 1.2 Osteoporosis & back pain & prostate (men) — — 11.7 (3) 7.3 (20) — — 1.3 1.2 Notes: O/E, observed/expected prevalence (see main text). The three most prevalent triads/highest O/E ranks in each year are printed in bold. Data source: the Spanish National Health Survey (ENSE). Int J Public Health | Owned by SSPH+ | Published by Frontiers 5 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity FIGURE 3 | PCA-derived components1 and chronic diseases with the highest factor loadings (A) and sex-specific, mean age and the proportion with less than primary school education of the total multimorbidity sample, the non-multimorbidity sample and those in the mms who are living with at least one chronic disease listed under each component (B), Spain 2006 vs. 20173. Note: 1. The order of the components are shown in brackets. No chronic disease had a factor loading of ≥0.3 in more than one component. 2. All loadings for thyroid dysfunction in 2017 were <0.3. 3. Total summary statistics for the multimorbid (mms) and non-multimorbid (nmms) samples pertain to the weighted population aged 60–89. Data source: the Spanish National Health Survey (ENSE). suggesting that these chronic diseases are not independent of highest O/E ratios belong to the 100 most prevalent each other. Both O/E ratios also increased over the studied combinations among individuals living with period. For women, the observed risk of suffering multimorbidity. Conversely, all ten triadic combinations simultaneously from hypertension, cholesterol and diabetes is with highest O/E ratios include constipation as one of the 1.5 higher than the expected risk. Worth mentioning is that the chronic conditions, six include hemorrhoids and four involve majority of triads with a high O/E ratio consists of combinations gastric ulcers. The combination of haemorrhoids/migraine/ of chronic diseases with relatively low prevalence. In the constipation holds the highest O/E ratio of 5.8 2017 data, none of the ten triadic combinations with the (Supplementary Table S3). Int J Public Health | Owned by SSPH+ | Published by Frontiers 6 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity Principal Component and Cluster Analyses represented in component 2 and 5. Likewise, the next cluster Lastly, on the group of participants with multimorbidity (mm- (thyroid dysfunction and varicose veins) corresponded to sample) we performed PCA and cluster analyses on the 20 single component 7 and cluster 3 (osteoarthrosis, osteoarthritis and conditions to identify frequently co-occurring conditions. For chronic back) to component 1 of the PCA. Cluster 4 and PCA 2006, the PCA derived eight components with eigenvalues component 4 were identical (COPD, allergies and chronic skin exceeding 1, explaining almost half of the variation in chronic problems), while cluster 5 coincided with component 6 in the case conditions (Supplementary Figure S2 and Supplementary of hypertension, diabetes and obesity but also included Table S4). The first component exhibited particularly high cholesterol and heart disease. Obesity was the last chronic factor loadings for osteoarthrosis and chronic back pain and, condition added to cluster 5, but given the commonality (all to a lesser extent, osteoporosis. Most other components also are cardiovascular disease risk factors), the cluster made sense contained closely-related chronic conditions with high factor from an epidemiological perspective. The same applied to the last loadings. For instance, component 3 contained urinary cluster (urinary incontinence and prostate problems). Turning to incontinence and prostate problems, component 4 COPD/ the 2017 results, not all clusters remained identical: obesity did asthma and allergies (and to a lesser extent chronic skin not join hypertension, diabetes and cholesterol, while the cluster problems) and component 5 haemorrhoids and constipation. with urinary incontinence and prostate problems included heart Regarding 2017, only six components had eigenvalues exceeding disease (which make clinical sense for men as these three chronic 1, which collectively explained 40.4% of the variance. However, diseases had an O/E of 1.9; results available upon request). Lastly, these components exhibited substantial overlap with the situation the first cluster included both chronic intestinal problems in 2006 regarding the most important chronic conditions they (haemorroids, constipation, gastric and duodenal ulcer, represented. As shown in Figure 3, the first and fourth varicose veins), metabolic issues (thyroid dysfunction) as well components for both years contain the same three chronic as mental health problems and migraine, almost coinciding with conditions that were selected for each component based on PCA components 3 and 7). their high factor loadings (>0.3), as indicated in Supplementary Table S4. Component 2 in 2017 has the same bowel-related issues and mental health conditions as in 2006, but DISCUSSION with the addition of haemorrhoids and constipation. The third component includes hypertension, cholesterol and diabetes, In this study, we analysed the prevalence of multimorbidity conditions that appeared less important among multimorbid (defined as suffering from at least three chronic conditions) individuals in 2006. among the 60–89 years-old population in Spain in 2006 and To determine sex, age and educational differences in the 2017, focusing on 20 chronic conditions. Overall, we found that disease patterns of the population living with multimorbidity, multimorbidity did not increase, as it affected six out of ten older the table in Figure 3 shows the characteristics of those who people in both years. However, there were some notable trends: manage one or more of the chronic diseases listed under each prevalence increased among men after age 64 and among women component. For instance, individuals suffering from after age 70. Importantly, the average number of conditions osteoporosis, osteoarthrosis and/or chronic back pain (listed among older adults with multimorbidity did not increase. under component 1) have similar demographic characteristics Regarding sex differences, in both 2006 and 2017, women with as the general population living with multimorbidity aged multimorbidity encountered, on average, over half a chronic 60–89 years. Conversely, those with mental health issues are, condition more than men. However, while the prevalence of on average, up to 1 year younger (statistically significant at the 5% multimorbidity among men increased by three percentage points, level for men in 2006), while people living with multimorbidity it decreased among women by two percentage points. and managing urinary incontinence and prostate problems Educational differences followed the expected pattern, with (component 3 in 2006) are 2–3 years older. Regarding changes higher-educated individuals having fewer conditions than the over time, the mean age of the male sample living with lower educated. Nevertheless, educational differences in the multimorbidity increased by 0.9 years, and for women with presence of multimorbidity widened over time. The prevalence multimorbidity by 1.7 years. Concurrently, the male and of multimorbidity was highest for those suffering from female populations with fewer than three chronic conditions hypertension, osteoarthrosis and chronic back pain (63.4%, were, on average, 0.8 and 0.9 years younger in 2017 compared 62.4%, and 55.9%, respectively, in 2017). However, the relative to 2006. This suggests that not only is multimorbidity not being risk for multimorbidity was highest for digestive-related delayed but that over the course of the decade, increases in life conditions, especially constipation and urinary incontinence expectancy were accompanied by increases in the time lived with (14.1 and 12.6, respectively, in 2017). multimorbidity. In 2017, the most common triadic combinations of chronic As for the cluster analyses we conducted (Supplementary conditions among multimorbid individuals included Table S5 and Supplementary Figure S3), the results closely osteoarthrosis, hypertension and/or back pain. However, when mirrored the components derived from the PCA. In the it comes to men, they also experienced increases in cholesterol dendrogram for the 2006 data, the first cluster included and diabetes, alongside a slight decline in chronic prostate haemorrhoids, constipation, gastric duodenal ulcer, mental problems between 2006 and 2017. For instance, the triad of health problems and migraine‒the main chronic conditions osteoporosis/back pain/prostate was ranked third in 2006 but Int J Public Health | Owned by SSPH+ | Published by Frontiers 7 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity 20th in 2017. Concurrently, the number of triadic combinations Our study has also revealed noteworthy and novel findings of chronic conditions that included prostate problems within the when comparing different periods and analysing sex and top 100 dropped from 30 to 21 during the same period. Possible educational differences. The observed sex gaps reinforce reasons include both better treatment and fewer diagnoses of previous literature on the poorer health experienced by prostate disorders due to new recommendations in prostate women. However, there are indications of a slight cancer screening in older men [39, 40]. Among multimorbid improvement among women, marked by a 2-percentage-point women, triadic combinations including osteoarthrosis, back pain reduction in the prevalence of multimorbidity, while men and/or hypertension were most prevalent, with little change over experienced a 3-percentage-point increase. This warrants time. It is also worth noting that the most frequent triadic closer examination in future research. Recognising educational combinations closely align with what we would anticipate inequalities in multimorbidity is important, as socioeconomic from the prevalence of individual conditions (i.e., the O/E differences often result in deprived populations, particularly ratios are approximately 1). Conversely, combinations that those with lower educational attainment, experiencing greater show interdependence (i.e., high O/E ratios) exhibit relatively health complexities. This underscores the significance of low prevalence, a trend resembling the findings in Germany [35] addressing and reducing social inequalities in health and concerning triadic combinations among health insurance policy identifying groups that require special attention from public holders aged 65 years and older. health services. For instance, a recent study conducted in the Comparing our findings with other studies directly is United Kingdom, which analysed various determinants of challenging due to a multitude of factors impacting the multimorbidity, indicated that the elevated risk of results, including the number and types of diseases multimorbidity among lower-educated individuals could be considered, sample demographics, underlying risk factors, partially attributed to their higher BMI and increased smoking and data collection methods. Nonetheless, by employing rates [45]. various methodological approaches [34] in which we have This study has certain limitations related to the data and demonstrated the importance of similar combinations of methodology employed in the analysis. Firstly, because it is a chronic conditions in different ways, our study still offers cross-sectional analysis, individuals cannot be followed over time, valuable insights for policymakers. Notably, two thus precluding discussions about individual changes in multimorbidity patterns identified in 2006 remained multimorbidity prevalence. Secondly, it is important to unchanged in 2017 according to both the PCA and cluster acknowledge that changes in disease prevalence may also analyses: 1) musculoskeletal disorders (osteoarthrosis, result from improvements in diagnostic practices and osteoarthritis and chronic back pain), which tend to alterations in medical protocols. Notably, revisions in increase with age [41]; and 2) environment-related chronic recommendations for low-density lipoprotein (LDL) levels in diseases (COPD, allergies and chronic skin problems), known 2011 [46] and 2016 [47] contributed to increased diagnosis of to be linked to increased air pollution resulting from rapid high cholesterol. Additionally, the Spanish Health Surveys only urbanization and industrialization [42]. Regarding the include individuals living in private households, excluding remaining chronic conditions, two other multimorbidity institutionalised individuals with likely poorer health statuses. patterns can be identified based on similar cluster and PCA While this exclusion could impact multimorbidity prevalence outcomes: 3) cardiovascular disease and associated risk factors: rates at older ages, it is worth noting that Spain’s institutional Hypertension, diabetes, obesity, cholesterol and heart disease arrangements encompass a relatively low proportion of the were one cluster and confined to two PCA components in population. In 2011, only 7% of individuals aged 85–89 years 2006. In 2017 they were more dispersed, with the were institutionalised, with even lower rates in younger age cardiometabolic risk factors hypertension, diabetes and brackets [48]. cholesterol all in one cluster and one PCA component. The Studying changes in the prevalence and patterns of chronic same applies to obesity, while heart disease formed a cluster diseases linked to multimorbidity is important from a and PCA component with prostate problems and urinary healthcare perspective. This is particularly relevant as health incontinence (known correlates of cardiovascular disease in systems remain largely single-illness oriented, despite the elderly men [43]). 4) Brain-gut comorbidity (one cluster in complex needs of patients with multimorbidity that are often both years, but two PCA components in 2006 and one in 2017), not addressed [34]. Furthermore, the persistence of disease containing gastrointestinal problems (haemorrhoids, clusters within multimorbidity patterns observed over time constipation, gastric duodenal ulcer), together with mental underscores the significance of addressing diseases through a health problems and migraine. These findings align with holistic approach rather than treating each chronic condition in previous research indicating a connection between isolation. Such an approach can prove highly effective in health gastrointestinal disorders and psychological symptoms [44]. prevention, enabling us to anticipate and mitigate future related Also noteworthy is that all ten triadic combinations with the issues. Considering the increasing number of people highest O/E ratios include constipation, six include experiencing health problems, driven in part by the hemorrhoids and four involve gastric ulcers. This suggests babyboom generation entering old age, the pressure on the that individuals with health issues related to the digestive healthcare system is expected to intensify in the coming decades. system are susceptible to experiencing various related Therefore, gaining a comprehensive understanding of change in chronic conditions simultaneously. disease prevalence, multimorbidity and the sex and Int J Public Health | Owned by SSPH+ | Published by Frontiers 8 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity socioeconomic changes therein is crucial for future healthcare ministry, a project on sex and socioeconomic inequalities in planning in an ageing population. comorbidity and multiple causes of death that is led by both researchers. The research was also supported by the European Research Council (ERC-2019-COG-864616, HEALIN) and the ETHICS STATEMENT CERCA Programme (Generalitat de Catalunya). The funders had no role in study design, data analysis, decision to publish, or Publically available secondary data were used. The studies were preparation of the manuscript. A previous version of this article conducted in accordance with the local legislation and was presented at the 2022 European Population Conference in institutional requirements. Written informed consent for Groningen, the Netherlands. participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements. CONFLICT OF INTEREST The authors declare that they do not have any conflicts of AUTHOR CONTRIBUTIONS interest. JS designed the study and obtained harmonised and analysed the data. JS and ER interpreted the results. JS wrote the first ACKNOWLEDGMENTS draft that was critically revised by ER. Both authors contributed to the article and approved the submitted version. We would like to thank Pilar Zueras for her useful comments and suggestions on a previous draft of this article. FUNDING SUPPLEMENTARY MATERIAL Both authors received funding from the Spanish Ministry of Science and Innovation through the Ramón y Cajal programme The Supplementary Material for this article can be found online at: (RYC-2013-14851 and RYC-2017-22586) and the R&D project https://www.ssph-journal.org/articles/10.3389/ijph.2023.1606259/ COMORHEALTHSES (PID2020-113934RB-I00) from the same full#supplementary-material Expectancy Among Elderly People in Spain: 1986–1999. Journals Gerontol REFERENCES Ser A: Biol Sci Med Sci (2005) 60(8):1028–34. doi:10.1093/gerona/60.8.1028 11. Zunzunegui MV, Nunez O, Durban M, García de Yébenes M-J, Béland F, 1. UNDESA (United Nations Department of Economic and Social Affairs). Otero Á. Decreasing Prevalence of Disability in Activities of Daily Living, 2022 Revision of World Population Prospects. New York (2022). Available Functional Limitations and Poor Self-Rated Health: A 6-Year Follow-Up Study From: https://population.un.org/wpp/ (Accessed August 29, 2023). in Spain. Aging Clin Exp Res (2006) 18(5):352–358. doi:10.1007/BF03324830 2. Omran AR. The Epidemiologic Transition Theory. A Preliminary Update. 12. Otero Á, Zunzunegui MV, Rodríguez-Laso Á, Aguilar MD, Lázaro P. Burden J Trop Pediatr (1983) 29:305–16. doi:10.1093/tropej/29.6.305 and Trends of Activities of Daily Living Disability in Elderly Spanish People. 3. Crimmins EM, Zhang Y, Saito Y. Trends Over 4 Decades in Disability-Free Rev Esp Salud Publica (2004) 78(2):201–13. doi:10.1590/s1135- Life Expectancy in the United States. Am J Public Health (2016) 106(7): 57272004000200007 1287–93. doi:10.2105/AJPH.2016.303120 13. Walter S, Beltrán-Sánchez H, Regidor E, Gomez-Martin C, del-Barrio JL, Gil- 4. Fries JF. Aging, Natural Death, and the Compression of Morbidity. New Engl de-Miguel A, et al. No Evidence of Morbidity Compression in Spain: A Time J Med (1980) 303(3):130–5. doi:10.1056/NEJM198007173030304 Series Study Based on National Hospitalization Records. Int J Public Health 5. Gispert R, Ruíz-Ramos M, Arán Barés M, Viciana F, Clot-Razquin G. (2016) 61(7):729–38. doi:10.1007/s00038-016-0829-5 Differences in Disability-Free Life Expectancy by Gender and Autonomous 14. Zueras P, Rentería E. Trends in Disease-Free Life Expectancy at Age 65 in Regions in Spain. Rev Esp Salud Publica (2007) 81(2):155–65. doi:10.1590/ Spain: Diverging Patterns by Sex, Region and Disease. PLOS ONE (2020) s1135-57272007000200006 15(11):e0240923. doi:10.1371/journal.pone.0240923 6. Jagger C, Robine J-M. Healthy Life Expectancy. In: Rogers RG, Crimmins EM, 15. Spijker J, Blanes-Llorens A. Mortality in Catalonia in the Context of the Third, editors. International Handbook of Adult Mortality, International Handbooks Fourth and Future Phases of the Epidemiological Transition Theory. Demogr of Population. Berlin: Springer (2011). p. 551–68. Res (2009) 20(8):129–68. doi:10.4054/demres.2009.20.8 7. Manton KG. Changing Concepts of Morbidity and Mortality in the Elderly 16. Solé-Auró A, Martín U, Domínguez Rodríguez A. Educational Inequalities in Population. Milbank Memorial Fund Q Health Soc (1982) 60:183–244. doi:10. Life and Healthy Life Expectancies Among the 50-Plus in Spain. Int J Environ 2307/3349767 Res Public Health (2020) 17(10):3558. doi:10.3390/ijerph17103558 8. Olshansky SJ, Rudberg MA, Carnes BA, Cassel CK, Brody JA. Trading Off 17. Lehnert T, Heider D, Leicht H, Heinrich S, Corrieri S, Luppa M, et al. Review: Longer Life for Worsening Health: The Expansion of Morbidity Hypothesis. Health Care Utilization and Costs of Elderly Persons With Multiple Chronic J Aging Health (1991) 3(2):194–216. doi:10.1177/089826439100300205 Conditions. Med Care Res Rev (2011) 68(4):387–420. doi:10.1177/ 9. Robine J-M, Saito Y, Jagger C. The Relationship Between Longevity and 1077558711399580 Healthy Life Expectancy. Qual Ageing Old Adults (2009) 10(2):5–14. doi:10. 18. Palladino R, Tayu Lee J, Ashworth M, Triassi M, Millett C. Associations 1108/14717794200900012 Between Multimorbidity, Healthcare Utilisation and Health Status: Evidence 10. Sagardui-Villamor J, Guallar-Castillón P, García-Ferruelo M, Banegas JR, From 16 European Countries. Age and Ageing (2016) 45(3):431–5. doi:10. Rodríguez-Artalejo F. Trends in Disability and Disability-Free Life 1093/ageing/afw044 Int J Public Health | Owned by SSPH+ | Published by Frontiers 9 October 2023 | Volume 68 | Article 1606259 Spijker and Rentería Shifting Patterns of Multimorbidity 19. Garin N, Olaya B, Moneta MV, Miret M, Lobo A, Ayuso-Mateos JL, et al. 35. Van den Bussche H, Koller D, Kolonko T, Hansen H, Wegscheider K, Glaeske Impact of Multimorbidity on Disability and Quality of Life in the Spanish G, et al. Which Chronic Diseases and Disease Combinations Are Specific to Older Population. PLOS ONE (2014) 9(11):e111498. doi:10.1371/journal.pone. Multimorbidity in the Elderly? Results of a Claims Data Based Cross-Sectional 0111498 Study in Germany. BMC public health (2011) 11(1):101–9. doi:10.1186/1471- 20. Orueta JF, García-Álvarez A, García-Goñi M, Paolucci F, Nuño-Solinís R. 2458-11-101 Prevalence and Costs of Multimorbidity by Deprivation Levels in the Basque 36. Fortin M, Stewart M, Poitras M-E, Almirall J, Maddocks H. A Systematic Country: A Population Based Study Using Health Administrative Databases. Review of Prevalence Studies on Multimorbidity: Toward a More PLoS One (2014) 9(2):e89787. doi:10.1371/journal.pone.0089787 Uniform Methodology. Ann Fam Med (2012) 10(2):142–51. doi:10. 21. Colombo F, García-Goñi M, Schwierz C. Addressing Multimorbidity to 1370/afm.1337 Improve Healthcare and Economic Sustainability. J Comorb (2016) 6(1): 37. De Vaus D. Surveys in Social Science. 5th ed. London: Routledge (2002). 21–7. doi:10.15256/joc.2016.6.74 38. Cornell JE, Pugh JA, Williams JW, Jr, Kazis L, Lee AF, Parchman ML, et al. 22. Salive ME. Multimorbidity in Older Adults. Epidemiologic Rev (2013) 35(1): Multimorbidity Clusters: Clustering Binary Data From Multimorbidity 75–83. doi:10.1093/epirev/mxs009 Clusters: Clustering Binary Data From a Large Administrative Medical 23. Foguet-Boreu Q, Violan C, Roso-Llorach A, Rodriguez-Blanco T, Pons-Vigués Database. Appl multivariate Res (2007) 12(3):163–82. doi:10.22329/amr. M, Muñoz-Pérez MA, et al. Impact of Multimorbidity: Acute Morbidity, Area v12i3.658 of Residency and Use of Health Services Across the Life Span in a Region of 39. Cayuela L, Lendínez-Cano G, Chávez-Conde M, Rodríguez-Domínguez S, South Europe. BMC Fam Pract (2014) 15(1):55. doi:10.1186/1471-2296-15-55 Cayuela A. Recent Trends in Prostate Cancer in Spain. Actas Urológicas 24. Carrasco-Ribelles LA, Roso-Llorach A, Cabrera-Bean M, Costa-Garrido A, Españolas (English Edition) (2020) 44(7):483–8. doi:10.1016/j.acuro.2020. Zabaleta-Del-Olmo E, Toran-Monserrat P, et al. Dynamics of Multimorbidity 05.003 and Frailty, and Their Contribution to Mortality, Nursing Home and Home 40. Hoffman RM. Striking the Right Balance With Prostate Cancer Screening. Care Need: A Primary Care Cohort of 1 456 052 Ageing People. JAMA Netw open (2022) 5(3):e222174–e74. doi:10.1001/jamanetworkopen. EClinicalMedicine (2022) 52:101610. doi:10.1016/j.eclinm.2022.101610 2022.2174 25. Prados-Torres A, Poblador-Plou B, Gimeno-Miguel A, Calderón-Larrañaga A, 41. Gheno R, Cepparo JM, Rosca CE, Cotten A. Musculoskeletal Disorders in the Poncel-Falcó A, Gimeno-Feliú LA, et al. Cohort Profile: The Epidemiology of Elderly. J Clin Imaging Sci (2012) 2(3):39–8. doi:10.4103/2156-7514.99151 Chronic Diseases and Multimorbidity. The EpiChron Cohort Study. Int 42. Baldacci S, Maio S, Cerrai S, Sarno G, Baïz N, Simoni M, et al. Allergy and J Epidemiol (2018) 47(2):382–4f. doi:10.1093/ije/dyx259 Asthma: Effects of the Exposure to Particulate Matter and Biological Allergens. 26. Olaya B, Moneta MV, Caballero FF, Tyrovolas S, Bayes I, Ayuso-Mateos JL, Respir Med (2015) 109(9):1089–104. doi:10.1016/j.rmed.2015.05.017 et al. Latent Class Analysis of Multimorbidity Patterns and Associated 43. Sandfeldt L, Hahn RG. Cardiovascular Risk Factors Correlate With Prostate Outcomes in Spanish Older Adults: A Prospective Cohort Study. BMC Size in Men With Bladder Outlet Obstruction. BJU Int (2003) 92(1):64–8. Geriatr (2017) 17(1):186. doi:10.1186/s12877-017-0586-1 doi:10.1046/j.1464-410X.2003.04277.x 27. Garin N, Olaya B, Perales J, Moneta MV, Miret M, Ayuso-Mateos JL, et al. 44. Tornkvist NT, Simrén M, Hreinsson JP, Tack J, Bangdiwala SI, Sperber AD, Multimorbidity Patterns in a National Representative Sample of the Spanish Adult et al. Prevalence and Impact of Disorders of Gut–Brain Interaction in Sweden. Population. PLOS ONE (2014) 9(1):e84794. doi:10.1371/journal.pone.0084794 Neurogastroenterology Motil (2023) 35:e14578. doi:10.1111/nmo.14578 28. Garin N, Koyanagi A, Chatterji S, Tyrovolas S, Olaya B, Leonardi M, et al. 45. North T-L, Harrison S, Bishop DC, Wootton RE, Carter AR, Richardson TG, Global Multimorbidity Patterns: A Cross-Sectional, Population-Based, Multi- et al. Educational Inequality in Multimorbidity: Causality and Causal Country Study. J Gerontol A Biol Sci Med Sci (2016) 71(2):205–14. doi:10.1093/ Pathways. A Mendelian Randomisation Study in UK Biobank. BMC public gerona/glv128 health (2023) 23(1):1644. doi:10.1186/s12889-023-16369-1 29. Forjaz MJ, Rodriguez-Blazquez C, Ayala A, Rodriguez-Rodriguez V, de Pedro- 46. Reiner Z, Catapano AL, De Backer G, Graham I, Taskinen MR, Wiklund O, Cuesta J, Garcia-Gutierrez S, et al. Chronic Conditions, Disability, and Quality et al. ESC/EAS Guidelines for the Management of Dyslipidaemias: The Task of Life in Older Adults With Multimorbidity in Spain. Eur J Intern Med (2015) Force for the Management of Dyslipidaemias of the European Society of 26(3):176–81. doi:10.1016/j.ejim.2015.02.016 Cardiology (ESC) and the European Atherosclerosis Society (EAS). Eur Heart J 30. He Y, Zhang M, Lin E, Bruffaerts R, Posada-Villa J, Angermeyer M, et al. (2011) 32(14):1769–818. doi:10.1093/eurheartj/ehr158 Mental Disorders Among Persons With Arthritis: Results From the World 47. Lloyd-Jones DM, Morris PB, Ballantyne CM, Birtcher KK, Daly DD, DePalma Mental Health Surveys. Psychol Med (2008) 38(11):1639–50. doi:10.1017/ SM, et al. 2016 ACC Expert Consensus Decision Pathway on the Role of Non- S0033291707002474 Statin Therapies for LDL-Cholesterol Lowering in the Management of 31. Marengoni A, Angleman S, Melis R, Mangialasche F, Karp A, Garmen A, et al. Atherosclerotic Cardiovascular Disease Risk: A Report of the American Aging With Multimorbidity: A Systematic Review of the Literature. Ageing Res College of Cardiology Task Force on Clinical Expert Consensus Rev (2011) 10(4):430–9. doi:10.1016/j.arr.2011.03.003 Documents. J Am Coll Cardiol (2016) 68(1):92–125. doi:10.1016/j.jacc.2016. 32. St John PD, Tyas SL, Menec V, Tate R. Multimorbidity, Disability, and Mortality in 03.519 Community-Dwelling Older Adults. Can Fam Physician (2014) 60(5):e272–e280. 48. Eurostat. The Eurostat Census Hub (2023). Available From: https://ec.europa. 33. Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology eu/CensusHub2 (Accessed April 30, 2023). of Multimorbidity and Implications for Health Care, Research, and Medical Education: A Cross-Sectional Study. Lancet (2012) 380(9836):37–43. doi:10. Copyright © 2023 Spijker and Rentería. This is an open-access article distributed 1016/S0140-6736(12)60240-2 under the terms of the Creative Commons Attribution License (CC BY). The use, 34. Islam MM, Valderas JM, Yen L, Dawda P, Jowsey T, McRae IS. Multimorbidity distribution or reproduction in other forums is permitted, provided the original and Comorbidity of Chronic Diseases Among the Senior Australians: author(s) and the copyright owner(s) are credited and that the original publication Prevalence and Patterns. PloS one (2014) 9(1):e83783. doi:10.1371/journal. in this journal is cited, in accordance with accepted academic practice. No use, pone.0083783 distribution or reproduction is permitted which does not comply with these terms. Int J Public Health | Owned by SSPH+ | Published by Frontiers 10 October 2023 | Volume 68 | Article 1606259