The Disabling Effect of Diseases: A Study on Trends in Diseases, Activity Limitations, and Their Interrelationships

Abstract

Objectives. Data from the Netherlands indicate a recent increase in prevalence of chronic diseases and a stable prevalence of disability, suggesting that diseases have become less disabling. We studied the association between chronic diseases and activity limitations in the Netherlands from 1990 to 2008.

Methods. Five surveys among noninstitutionalized persons aged 55 to 84 years (n = 54 847) obtained self-reported data on chronic diseases (diabetes, heart disease, peripheral arterial disease, stroke, lung disease, joint disease, back problems, and cancer) and activity limitations (Organisation for Economic Co-operation and Development [OECD] long-term disability questionnaire or 36-item Short Form Health Survey [SF-36]).

Results. Prevalence rates of chronic diseases increased over time, whereas prevalence rates of activity limitations were stable (OECD) or slightly decreased (SF-36). Associations between chronic diseases and activity limitations were also stable (OECD) or slightly decreased (SF-36). Surveys varied widely with regard to disease and limitation prevalence rates and the associations between them.

Conclusions. The hypothesis that diseases became less disabling from 1990 to 2008 was only supported by results based on activity limitation data as assessed with the SF-36. Further research on how diseases and disability are associated over time is needed.

As populations in many Western countries grow older, changes in population health constitute a major public health issue. Insight into health-status trends in older populations may help with estimating future health care needs and setting priorities for improving population health. Important indicators for health in old age are prevalence of chronic diseases and prevalence of disability.

Previous studies in the Netherlands showed that the prevalence of chronic diseases increased over the past 20 to 25 years.^1,2 This increase was not only attributable to aging of the population; age-group-specific prevalence rates also increased.³ In contrast, the age-adjusted prevalences of activity limitations and disabilities have shown varying trends in the Netherlands: rates have remained constant, increased, or even decreased, depending on the study.^4–7 Decreasing disability trends have also been observed in several other Western countries, suggesting a postponement of limitations and disabilities despite increases in chronic diseases in many countries.⁸ Because diseases are important causes of activity limitations and disabilities,^9,10 these opposing trends warrant further study.

Several possible explanations for these contradictory trends have been suggested. First, trends in disability rates do not necessarily follow trends in chronic diseases, because increasing use of aids and devices facilitates greater independence in people with or without diseases.^11,12 Also, changes in the environment can mitigate disabilities for people with a disease.¹³ Safer and better sidewalks, for example, can make it easier for people with joint disease to move around. Another explanation is that improved medical knowledge and health services utilization lead to more detection of disease even though the actual disease prevalence remains the same. For some diseases, such as type 2 diabetes, hypertension, and some cancers, people are now being diagnosed earlier and are receiving better treatment than before.¹¹ These advancements in detection and treatment lead to longer periods of known morbidity—and thus to higher estimates of disease prevalence—but with improved functional status. These trends would result in stable or decreasing prevalence of activity limitations and disabilities.^8,13,14 In that case, the conclusion would be that diseases have become less disabling over time. For different diseases, different underlying developments are possible, resulting in different trends in the disabling impact of diseases.⁶

We sought to investigate time trends in the disabling impact of chronic diseases in the Netherlands. We examined time trends in disease prevalence, activity limitations, and the strength of the associations between diseases and activity limitations. To ensure the best estimates of these trends, we combined all available data sources.

METHODS

In the Netherlands, several surveys elicit information on diseases and functional limitations, and no single survey can be regarded as the best. Therefore, we selected all available Dutch population-based surveys relevant to the subject. Our inclusion criteria for the studies were that they should

1. contain data on self-reported activity limitations and chronic diseases,

2. have at least 3 data-collection moments,

3. represent both genders,

4. cover a minimum period of 10 years, and

5. survey noninstitutionalized respondents aged 55 to 84 years.

We harmonized original data from 5 surveys: the Amenities and Services Utilization Survey (Dutch abbreviation: AVO),¹⁵ the Netherlands Health Interview Survey (renamed Permanent Life Situation Survey [Dutch abbreviation: POLS] in 1997),¹⁶ the Longitudinal Aging Study Amsterdam (LASA),¹⁷ the Maastricht Aging Study (MAAS),¹⁸ and the Doetinchem Cohort Study (DCS).¹⁹ The surveys differ in a number of aspects, including whether proxy measurements were used, frequency of measurement, whether the survey covers a region or the whole country, and whether it is a repeated cross-sectional survey or a time-series survey. Table 1 gives information on these and other general characteristics of the 5 surveys. Many of these characteristics cannot be adjusted for in the models. We therefore analyzed all surveys separately so we could combine the results in a meta-analysis.

TABLE 1—

General Characteristics and Prevalence Rates of Diseases and Activity Limitations of the Selected Surveys: The Netherlands, 1990–2008

Study Characteristics and Measures	AVO	POLS	LASA	MAAS	DCS
Commissioning organization	Netherlands Institute for Social Research	Statistics Netherlands	VU University Amsterdam	Maastricht University	National Institute for Public Health and the Environment
Geographical representation	National	National	Multiregional	Provincial	Municipal
Type of survey	Repeated cross-sectional	Repeated cross-sectional	Longitudinal	Longitudinal	Longitudinal
Frequency of measurement, y	4	1	3	3	5
Duration of 1 cycle of measurement, y	1	1	2a	3	5
Study period	1991–2007	1990–2008	1995–2006a	1993–2007	1995–2007b
Interview vs questionnaire	Interview and questionnaire	Interview and questionnaire	Interview and questionnaire	Questionnaire	Questionnaire
Use of proxies	Allowed	Allowed	Allowed	Not allowed	Not allowed
Baseline age range, y	≥ 6	≥ 0	55–84	24–81	20–59
Baseline age range of respondents  selected for the current study, y	55–84	55–84	55–84	55–81	55–59
Response range, % of households	43–70	60–65	62–82	54–75	62–80
Selected respondents,c no.	3411	2190	1638	934	2037
Diseases, %d
Diabetes	8.8	8.3	9.1	7.4	4.8
Heart disease	9.6	9.2	23.8	16.2	4.2
Peripheral arterial disease	8.0	5.3	10.1	6.2	24.1
Stroke	3.3	3.4	5.8	NA	2.5
Lung disease	10.5	9.8	13.8	11.3	3.6
Joint disease	25.0	26.5	47.6	17.4	NA
Back problems	15.6	13.6	NA	NA	59.4
Cancer	3.2	6.6	13.4	7.7	7.1
Morbidity summary, %
Exactly 1 disease	24.1	29.4	37.3	27.3	43.3
≥ 2 diseases	20.7	23.9	34.3	14.8	27.8
Activity limitations, %b
At least moderate OECD	24.0	32.2	40.3	NA	NA
Severe OECD	4.8	7.8	13.8	NA	NA
At least moderate SF-36	NA	NA	NA	52.1	38.0
Severe SF-36	NA	NA	NA	16.8	9.4

Note. AVO = Amenities and Services Utilization Survey; DCS = Doetinchem Cohort Study; LASA = Longitudinal Aging Study Amsterdam; MAAS = Maastricht Aging Study; NA = not available; OECD = Organisation for Economic Co-operation and Development long-term disability questionnaire; POLS = Permanent Life Situation Survey; SF-36 = 36-item Short Form Health Survey.

^aExcept for the period 2001–2003, when the old cohort was interviewed in 2001–2002 and a new cohort was interviewed in 2002–2003. In 2005–2006 both cohorts were interviewed at the same time.

^bWaves in 1993 and 1994 did not include 36-item Short Form Health Survey data and were excluded from the analyses.

^cAveraged over all years of measurement.

^dCrude prevalences; mean over all years of measurement.

Respondents who reached age 55 years between cycles of data collection were added to the data set in the subsequent cycle. Also, in 2003 a new cohort aged 55 to 65 years was added to the surviving LASA cohort. These 2 groups of “new” respondents lacked self-reports in the years before their 55th birthday and before 2003, respectively.

Measurement of Diseases and Activity Limitations

On the basis of the condition that at least 3 surveys had to collect information on self-reported diseases in a comparable way (Table 2), we selected 8 diseases for analysis: diabetes, heart disease, peripheral arterial disease, stroke, lung disease, joint diseases, back problems, and cancer. We constructed 1 summary measure to distinguish people with morbidity (no disease vs ≥ 1 disease) and another to distinguish people with multimorbidity (no disease vs 1 disease vs ≥ 2 diseases).

TABLE 2—

Survey Questionnaire and Interview Wording on Selected Diseases: The Netherlands, 1990–2008

	AVO	POLS	LASA	MAAS	DCS
Introductory text	The following is a list of diseases and conditions. Please indicate whether you currently have any of these diseases or conditions or have had them in the past 12 mo.a	I am going to mention a number of diseases and conditions. Please indicate whether you currently have any of these or have had them in the past 12 mo.b	Now, I'd like to ask about a number of diseases. It's about diseases and symptoms that last for at least 3 mo or for which people need to be treated by a physician. Please indicate whether you currently have any of the diseases that I list.	Have you been told by a physician within the past 3 y that you suffer from any of the following conditions?	NA
Diabetes	Diabetes	Do you have diabetes?cd	Do you have diabetes?	Diabetes (IDD or NIDD)	Do you have diabetes?
Heart disease	Severe heart condition (such as heart failure or angina), (consequences of a) heart attacke	Did you ever have a heart attack? Did you have another severe heart condition in the past 12 mo?df	Do you have heart disease, or did you have a heart attack?	Heart failure or cardiac insufficiency; heart attack	Did you ever have a heart attack?
Peripheral arterial disease	Narrowing of veins in the abdomen or legs (no varicosis)g	Narrowing of veins in the abdomen or legs (no varicosis)h	Do you have abnormalities or diseases of arteries or veins in the abdomen or legs?	Peripheral arterial diseases or claudication	Do you experience pain in your legs when walking?
Stroke	(Consequences of) stroke, brain hemorrhage, or cerebral infarctioni	Did you ever have a stroke, brain hemorrhage, or cerebral infarction?dj	Did you ever have a stroke or a brain attack? (also brain hemorrhage)	NA	Did you ever have a stroke?
Lung disease	Asthma, chronic bronchitis, emphysema, or CNSLDk	Asthma, chronic bronchitis, emphysema, or CNSLDl	Do you have CNSLD (asthma, chronic bronchitis, emphysema)?	CNSLD, chronic bronchitis, or emphysema	Did you ever have asthma?
Joint disease	Arthritis of knees, hips; rheumatoid arthritism	Arthritis of knees, hips, or hands; rheumatoid arthritisn	Do you have arthritis of knees, hips, or hands? Do you (also) have rheumatoid arthritis?	Rheumatic condition or joint disease	NA
Back problems	Severe or long-lasting back problemso	Severe or long-lasting back problemsp	NA	NA	In the past 12 mo, did you have problems (pain, discomfort) in the low back? In the past 12 mo, did you have problems (pain, discomfort) in the upper back?
Cancer	Cancer or malignant conditionq	Did you ever have any kind of cancer (malignant condition)?dr	Do you have tumor formation or cancer, or did you have it in the past?	Cancer or malignancy	Do you have any kind of cancer, or did you have it in the past?

Note. AVO = Amenities and Services Utilization Survey; CNSLD = chronic nonspecific lung disease; DCS = Doetinchem Cohort Study; IDD = insulin-dependent diabetes; LASA = Longitudinal Aging Study Amsterdam; MAAS = Maastricht Aging Study; NA = not available; NIDD = non–insulin-dependent diabetes; POLS = Permanent Life Situation Survey;

^aWording used in 1995 differed from the wording in this column; no disease information available before 1995.

^bInterview wording used in the period 1990–2000 differed from the wording used in the questionnaire (in this column).

^cBefore 2001 this question read “diabetes.”

^dThese questions precede the introductory text.

^eTwo separate questions; in 1995 this question read “Severe heart condition or heart attack.”

^fTwo separate questions; before 2001 this question read “Severe heart condition or heart attack.”

^gNot available before 1999.

^hNot available before 2001.

ⁱIn 1995 this question read “(Consequences of a) stroke.”

^jBefore 2001 this question read “(Consequences of a) stroke.”

^kIn 1995 this question read “Asthma, chronic bronchitis, or CNSLD.”

^lBefore 2001 this question read “Asthma.”

^mTwo separate questions; in 1995 the first question read “Arthritis of knees, hips, or hands.”

ⁿTwo separate questions; before 2001 these were 3 questions that read “Arthritis of knees, hips, or hands; rheumatoid arthritis; other chronic rheumatoid condition lasting for 3 mo or more.”

^oIn 1995 this question read “Persistent back problems, lasting longer than 3 mo, or spinal hernia.”

^pBefore 2001 this question read “Persistent back problems, lasting longer than 3 mo.”

^qIn 1995 the order of cancer and malignant neoplasm was reversed.

^rBefore 2001 this question read “Malignant condition or cancer.”

We based the assessment of activity limitations on 3 items that the 5 surveys had in common: stair climbing, walking outside, and getting dressed.⁷ AVO, POLS, and LASA used variants of items from the Organisation for Economic Co-operation and Development (OECD) long-term disability questionnaire, and MAAS and DCS used items based on the Medical Outcomes Study 36-item Short Form Health Survey (SF-36) to assess activity limitations in these domains.^20–22 In our analyses, we differentiated between results derived from the OECD questionnaire and those derived from the SF-36 because of differences in their wording (“difficulty in performing actions” vs “health-limiting effects in the performance of activities,” respectively) and because earlier studies suggested that there were differences in the criterion validity of the 2 questionnaires.^7,23,24 We combined the 3 items to create 1 summary measure for activity limitations derived from the OECD and 1 derived from the SF-36. For both, we distinguished 3 levels of limitation: (1) not limited at all, (2) moderately limited in at least 1 activity, and (3) severely limited in at least 1 activity. We further dichotomized these categories into 2 variables indicating either severe limitations or at least moderate (including severe) limitations.

Statistical Analyses

We used descriptive analyses to present the prevalence of the selected diseases and activity limitations in the separate surveys averaged over all observation years. Then we used regression models to estimate time trends for (1) each disease and morbidity summary measure, (2) each activity limitation, and (3) the association between disease or morbidity summary measures and activity limitations. In models 1 and 2, we used year of observation as the independent variable. In model 3, the disease or morbidity summary variable, the year of observation, and the interactions between these 2 variables served as independent variables. We carried out the regression analyses on each of the 5 data sets and adjusted all parameters for age and gender. For the prospective studies, we fitted a generalized estimation equations model; for the cross-sectional studies, we fitted a generalized linear model.²⁵ More details on the regression models are given in Appendix A (available as a supplement to the online version of this article at http://www.ajph.org).

To reduce loss of information because of missing values in the data on diseases and activity limitations, we imputed missing values by using the multiple imputations by chained equations algorithm.²⁶ The proportion of data missing for most variables varied from 1% to 20%, with 2 exceptions: 56% missing for peripheral arterial disease in POLS because this disease was not measured before 2001; and 43% missing in SF-36 in MAAS. We used 10 imputations per survey including all data on diseases, activity limitations, and demographic characteristics (age, gender, marital status, and household size). Per survey, we pooled the regression parameters of the 10 imputations to gain survey-specific estimates by using Rubin's rules.²⁷

To get overall estimates we performed meta-analyses on all survey-specific estimates. Because there might be unexplained heterogeneity between surveys that could be attributed to survey-specific aspects, such as data-sampling methods and interviewing methods, we used a random-effects model based on the DerSimonian-Laird method.²⁸ The random-effects model assumes that each study has its own mean and (within-study) variance for its effect, and the mean of each study comes from a hyperdistribution of means. That distribution is assumed to be Gaussian, with an overall mean and an overall variance (between-study variance, which quantifies the amount of heterogeneity). Estimating the overall mean while taking into account the heterogeneity allows for inference over all studies.^29,30 Tests for heterogeneity justified the choice of the random-effects model. This will generally yield wider confidence intervals than fixed-effect models, resulting in a less likely rejection of the null hypothesis (i.e., results are less likely to be significant and are thus more conservative).

RESULTS

General characteristics of the 5 surveys are summarized in Table 1. Two of the 5 surveys had collected nationally representative data with a repeated cross-sectional design. The other 3 were longitudinal studies performed in 1 specific region or in several regions. Response rates varied from 43% to 82%. Mean age varied from 60 to 70 years.

Three surveys showed that about 50% (42% to 54%) of the respondents had at least 1 disease (Table 1). In the other 2 surveys, this rate was a little higher than 70%. Prevalence rates of stroke, diabetes, and lung disease were fairly consistent across the surveys, with the DCS study reporting the lowest prevalences. These observed differences might be explained by differences in mean age because the DCS study had the youngest study population. The differences in prevalence rates of the other diseases were greater. Methods for ascertaining the presence of these diseases might explain these differences (Table 2). For example, the prevalence of peripheral arterial disease was much higher in the DCS study than in the other surveys. This is most likely attributable to the nonspecific way this disease was described for respondents. The prevalence of at least moderate activity limitations varied from 24% to 52%, and the prevalence of severe limitations varied from 5% to 17%.

Trends in Chronic Diseases and Disabilities

In the period 1990 to 2008, the odds of reporting diabetes increased by 5% per year (odds ratio [OR] = 1.05; 95% confidence interval [CI] = 1.04, 1.07; Table 3). In addition, the odds of reporting stroke increased 5% per year (OR = 1.05; 95% CI = 1.01, 1.08), and the odds of reporting cancer increased 6% per year (OR = 1.06; 95% CI = 1.01, 1.12). We found stable prevalence rates for heart disease, peripheral arterial disease, lung disease, joint disease, and back problems. Overall, the odds of reporting 1 or more of the selected diseases increased by 3% per year (OR = 1.03; 95% CI = 1.00, 1.05), and the odds of reporting 2 or more diseases increased by 3% per year (OR = 1.03; 95% CI = 1.01, 1.04). For most diseases the results were fairly consistent across studies. The only exception was heart disease: 3 studies found an increasing trend, whereas the other 2 found this trend to be declining (although not significantly).

TABLE 3—

Trends in Prevalences of Chronic Diseases and Activity Limitations Among Persons Aged 55–84 Years: The Netherlands, 1990–2008

Trends	AVO, OR (95% CI)	POLS, OR (95% CI)	LASA, OR (95% CI)	MAAS, OR (95% CI)	DCS, OR (95% CI)	Overall, OR (95% CI)
Disease specific
Diabetes	1.07* (1.04, 1.09)	1.05* (1.04, 1.05)	1.07* (1.05, 1.09)	1.01 (0.97, 1.04)	1.06* (1.01, 1.10)	1.05* (1.04, 1.07)
Heart disease	1.04* (1.02, 1.06)	1.03* (1.03, 1.04)	1.02* (1.01, 1.04)	0.99 (0.96, 1.01)	0.96 (0.91, 1.01)	1.02 (1.00, 1.04)
Peripheral arterial disease	0.94* (0.90, 0.99)	1.01 (0.98, 1.04)	0.99 (0.97, 1.01)	1.00 (0.97, 1.04)	0.95* (0.93, 0.97)	0.98 (0.95, 1.00)
Stroke	1.05* (1.02, 1.08)	1.08* (1.07, 1.09)	1.03* (1.01, 1.06)	NA	1.00 (0.93, 1.07)	1.05* (1.01, 1.08)
Lung disease	1.06* (1.04, 1.08)	1.01* (1.00, 1.01)	1.02* (1.00, 1.04)	0.99 (0.96, 1.01)	0.99 (0.94, 1.04)	1.01 (0.99, 1.04)
Joint disease	0.95* (0.94, 0.97)	1.01* (1.01, 1.02)	1.03* (1.01, 1.04)	1.05* (1.03, 1.08)	NA	1.01 (0.98, 1.04)
Back problems	1.01 (1.00, 1.03)	1.01* (1.00, 1.02)	NA	NA	1.09* (1.07, 1.11)	1.04 (1.00, 1.08)
Cancer	1.07* (1.03, 1.10)	1.14* (1.12, 1.15)	1.04* (1.02, 1.06)	1.02 (0.99, 1.05)	1.05* (1.01, 1.09)	1.06* (1.01, 1.12)
Morbidity summary 1: ≥ 1 disease	0.98* (0.97, 1.00)	1.03* (1.02, 1.03)	1.04* (1.02, 1.05)	1.00 (0.98, 1.02)	1.09* (1.07, 1.11)	1.03* (1.00, 1.05)
Morbidity summary 2
Exactly 1 disease	0.96* (0.94, 0.97)	1.00 (0.99, 1.02)	1.00 (0.98, 1.01)	0.98 (0.96, 1.03)	1.08* (1.06, 1.11)	1.00 (0.97, 1.03)
Multimorbidity (≥ 2 diseases)	1.02* (1.00, 1.04)	1.04* (1.04, 1.05)	1.04* (1.03, 1.06)	1.03* (1.00, 1.05)	0.99 (0.97, 1.01)	1.03* (1.01, 1.04)
Activity limitations
≥ moderate OECD	1.02* (1.01, 1.03)	0.99* (0.99, 1.00)	1.05* (1.04, 1.06)	NA	NA	1.02 (0.99, 1.05)
Severe OECD	1.03* (1.01, 1.05)	0.99* (0.98, 0.99)	1.02 (1.00, 1.04)	NA	NA	1.01 (0.98, 1.04)
≥ moderate SF-36	NA	NA	NA	0.97 (0.95, 1.00)	0.95* (0.94, 0.97)	0.96* (0.94, 0.98)
Severe SF-36	NA	NA	NA	0.98* (0.95, 0.99)	0.97* (0.94, 0.99)	0.97* (0.95, 0.99)

Note. AVO = Amenities and Services Utilization Survey; CI = confidence interval; DCS = Doetinchem Cohort Study; LASA = Longitudinal Aging Study Amsterdam; MAAS = Maastricht Aging Study; NA = not available; OECD = Organisation for Economic Co-operation and Development long-term disability questionnaire; OR = odds ratio; POLS = Permanent Life Situation Survey; SF-36 = 36-item Short Form Health Survey.

^*P ≤ .05.

The odds of reporting at least moderate SF-36 limitations decreased by 4% per year from 1990 to 2008 (OR = 0.96; 95% CI = 0.94, 0.98), and the odds of reporting severe SF-36 limitations decreased by 3% per year (OR = 0.97; 95% CI = 0.95, 0.99; Table 3). In the same time period, overall trends in activity limitations as assessed with the OECD instrument were stable. However, results were not consistent among surveys. POLS, MAAS, and DCS showed decreasing trends in activity limitations, whereas AVO and LASA showed increasing trends.

Trends in Associations Between Diseases and Activity Limitations

There was a strong association between diseases and activity limitations. The surveys were fairly comparable in this regard, with joint disease, stroke, and heart disease showing consistently stronger associations with activity limitations (see Table A1 in the appendix, available as a supplement to the online version of this article at http://www.ajph.org).

For most diseases, we observed no statistically significant changes in the strength of the associations with activity limitations in the period 1990 to 2008 (Table 4). However, the association between stroke and severe OECD limitations decreased over time by 6% per year (OR = 0.94; 95% CI = 0.91, 0.97). Also, the association between peripheral arterial disease and at least moderate SF-36 limitations decreased by 5% per year (OR = 0.95; 95% CI = 0.91, 1.00). For all diseases taken together, the association with activity limitations as assessed with the OECD instrument was stable over time, and the association with at least moderate SF-36 activity limitations decreased by 6% per year (OR = 0.94; 95% CI = 0.89, 1.00; Table 4). Also, the association of having exactly 1 disease with at least moderate or severe SF-36 limitations decreased by 6% and 5%, respectively (OR = 0.94; 95% CI = 0.92, 0.97 and OR = 0.95; 95% CI = 0.92, 0.99, respectively). There were, however, remarkable differences between surveys: POLS—a large, repeated, cross-sectional survey that assessed limitations with the OECD—showed a significantly weakening association between diseases and limitations. This weakening association was seen for almost all diseases and for at least moderate limitations as well as for severe limitations. Both DCS and MAAS also found the associations between diseases and limitations becoming less strong over time, although not as convincingly as the POLS study. On the other hand, LASA and AVO showed the disabling effect of some diseases becoming stronger over time.

TABLE 4—

Trends in Relations Between Diseases and Activity Limitations Among Persons Aged 55 to 84 Years: The Netherlands, 1990–2008

	OECD Activity Limitations, OR (95% CI)								SF-36 Activity Limitations, OR (95% CI)
	At Least Moderate (AVO)	Severe (AVO)	At Least Moderate (POLS)	Severe (POLS)	At Least Moderate (LASA)	Severe (LASA)	At Least Moderate (Overall)	Severe (Overall)	At Least Moderate (MAAS)	Severe (MAAS)	At Least Moderate (DCS)	Severe (DCS)	At Least Moderate (Overall)	Severe (Overall)
Disease-specific
Diabetes	0.94 (0.88, 1.00)	0.98 (0.89, 1.09)	0.97* (0.96, 0.99)	0.98 (0.95, 1.01)	1.06* (1.01, 1.11)	1.01 (0.95, 1.09)	0.99 (0.93, 1.05)	0.98 (0.96, 1.01)	0.97 (0.90, 1.04)	0.98 (0.90, 1.08)	0.99 (0.90, 1.09)	1.07 (0.93, 1.23)	0.98 (0.92, 1.03)	1.01 (0.93, 1.09)
Heart disease	1.11* (1.03, 1.19)	1.08 (0.97, 1.21)	0.96* (0.94, 0.98)	0.95* (0.92, 0.97)	1.03 (0.99, 1.06)	0.99 (0.94, 1.04)	1.02 (0.95, 1.10)	0.99 (0.93, 1.04)	0.98 (0.93, 1.04)	0.97 (0.89, 1.04)	0.96 (0.87, 1.06)	1.07 (0.94, 1.21)	0.98 (0.93, 1.03)	1.00 (0.91, 1.10)
Peripheral  arterial disease	1.03 (0.95, 1.11)	1.09 (0.96, 1.24)	0.99 (0.96, 1.02)	0.99 (0.96, 1.02)	1.08* (1.02, 1.13)	1.01 (0.96, 1.08)	1.03 (0.97, 1.09)	1.00 (0.97, 1.04)	0.92 (0.84, 1.01)	0.91 (0.82, 1.02)	0.96 (0.92, 1.01)	1.04 (0.97, 1.11)	0.95* (0.91, 1.00)	0.98 (0.86, 1.11)
Stroke	1.01 (0.91, 1.12)	0.95 (0.84, 1.07)	0.92* (0.90, 0.95)	0.93* (0.90, 0.97)	0.98 (0.93, 1.05)	0.96 (0.90, 1.03)	0.96 (0.91, 1.02)	0.94* (0.91, 0.97)	NA	NA	0.97 (0.85, 1.11)	0.98 (0.82, 1.16)	NA	NA
Lung disease	1.09* (1.03, 1.16)	1.21* (1.10, 1.33)	0.99 (0.97, 1.01)	0.98 (0.96, 1.01)	1.03 (0.99, 1.08)	1.02 (0.97, 1.08)	1.03 (0.98, 1.09)	1.05 (0.96, 1.15)	0.93 (0.87, 1.00)	0.95 (0.88, 1.03)	1.03 (0.94, 1.13)	1.04 (0.91, 1.19)	0.98 (0.89, 1.08)	0.98 (0.90, 1.07)
Joint disease	1.03 (0.99, 1.08)	1.01 (0.94, 1.08)	0.97* (0.96, 0.98)	0.96* (0.94, 0.97)	1.05* (1.03, 1.08)	1.02 (0.98, 1.06)	1.02 (0.96, 1.08)	0.99 (0.94, 1.04)	0.98 (0.93, 1.03)	1.00 (0.93, 1.06)	NA	NA	NA	NA
Back problems	1.11* (1.06, 1.16)	1.07 (0.99, 1.15)	0.98* (0.96, 0.99)	0.99 (0.97, 1.02)	NA	NA	1.04 (0.92, 1.17)	1.02 (0.95, 1.09)	NA	NA	0.95* (0.92, 0.97)	0.97 (0.92, 1.03)	NA	NA
Cancer	0.93 (0.84, 1.03)	0.98 (0.83, 1.15)	0.96* (0.94, 0.99)	0.96* (0.93, 1.00)	1.03 (0.98, 1.08)	0.99 (0.94, 1.06)	0.98 (0.93, 1.03)	0.97 (0.94, 1.00)	0.97 (0.87, 1.07)	0.97 (0.86, 1.08)	1.02 (0.94, 1.10)	1.10 (0.96, 1.25)	1.00 (0.94, 1.06)	1.03 (0.90, 1.16)
Morbidity Summary 1: ≥ 1 disease	1.02* (1.00, 1.04)	1.03 (1.00, 1.06)	0.98* (0.97, 0.98)	0.97* (0.97, 0.98)	1.05* (1.03, 1.06)	1.02 (0.99, 1.04)	1.02 (0.97, 1.07)	1.00 (0.97, 1.04)	0.97 (0.94, 1.00)	0.99 (0.95, 1.02)	0.92* (0.90, 0.94)	0.94* (0.91, 0.96)	0.94* (0.89, 1.00)	0.96 (0.91, 1.01)
Morbidity summary 2
Exactly 1 disease	1.03 (1.00, 1.06)	1.07* (1.01, 1.15)	0.97* (0.97, 0.98)	0.96* (0.95, 0.98)	1.04* (1.01, 1.06)	1.01 (0.98, 1.05)	1.01 (0.96, 1.06)	1.01 (0.95, 1.07)	0.95* (0.91, 1.00)	0.97 (0.92, 1.02)	0.94* (0.92, 0.97)	0.94* (0.90, 0.99)	0.94* (0.92, 0.97)	0.95* (0.92, 0.99)
Multimorbidity (≥ 2 diseases)	0.99 (0.96, 1.02)	0.99 (0.95, 1.03)	0.96* (0.95, 0.97)	0.96* (0.95, 0.97)	1.05* (1.03, 1.07)	1.01 (0.98, 1.04)	1.00 (0.94, 1.06)	0.99 (0.95, 1.02)	0.99 (0.95, 1.04)	0.99 (0.94, 1.04)	0.90* (0.87, 0.93)	0.95* (0.91, 0.99)	0.94 (0.85, 1.04)	0.97 (0.93, 1.01)

Note. AVO = Amenities and Services Utilization Survey; CI = confidence interval; DCS = Doetinchem Cohort Study; LASA = Longitudinal Aging Study Amsterdam; MAAS = Maastricht Aging Study; NA = not available; OECD = Organisation for Economic Co-operation and Development long-term disability questionnaire; OR = odds ratio; POLS = Permanent Life Situation Survey; SF-36 = 36-item Short Form Health Survey.

^*P ≤ .05.

DISCUSSION

The main question of our study was whether diseases became less disabling over the period 1990 to 2008. Based on meta-analyses of 5 large-scale Dutch surveys, we found (1) increasing prevalence rates of 3 out of 8 chronic diseases and of morbidity measures, (2) stable (OECD items) or decreasing (SF-36 items) activity limitation prevalence rates, and (3) stable (OECD) or decreasing (SF-36) associations between chronic diseases and activity limitations. The hypothesis that diseases became less disabling over the period 1990 to 2008 was therefore only supported by results based on activity limitation as assessed with the SF-36.

Explanation of Trends in Diseases and Activity Limitations

Our finding of an increasing prevalence of chronic diseases was based on self-reports. Better knowledge of diseases or a changing willingness to report diseases might have caused an increase in self-reported disease while the actual prevalence did not change. However, registrations in general practice have also shown an increase in the prevalence of chronic diseases.⁶ Furthermore, studies in most other countries have reported evidence of an increasing prevalence of chronic diseases.^3,31–35 The major explanations for this trend are changes in risk factor distributions, earlier diagnosis, and better survival among those with chronic diseases.¹⁴

For activity limitations, we found a stable prevalence without a clear trend (or at most a slight increasing trend). This was discussed in our earlier article on trends in activity limitations,⁷ in which we studied trends in the prevalence of separate limitations in 3 activities: walking, climbing stairs, and dressing and undressing. These earlier results showed stable prevalence rates of these separate limitations as assessed with the SF-36, and a slight increase in the prevalence of at least moderate activity limitations in stair climbing and in getting dressed as assessed with OECD items. In the current article, in which activity limitations were defined on the basis of the 3 items combined, we found a stable prevalence for OECD limitations and a slight decrease in prevalence for SF-36 limitations.

The main result of our study concerns the change in the association between diseases and activity limitations over time. Our hypothesis—that the strength of this association would grow weaker over time—could only be partly confirmed. We found the disabling effect of having a disease to be relatively constant based on activity limitation data as assessed with the OECD instrument, but we found weakening associations between diseases and activity limitations as assessed with the SF-36. However, results were not consistent across all surveys. Data from POLS, DCS, and MAAS suggested that diseases became less disabling from 1990 to 2008, and data from AVO and LASA suggested the contrary. Taken together, the overall results show a fairly stable association.

There are few other studies on trends in the disabling impact of chronic diseases. A previous Dutch study showed different results for different diseases, in the sense that the disabling impact of the more fatal diseases decreased, whereas the impact of nonfatal diseases increased.⁶ We did not replicate this finding in our current meta-analysis. Other recent studies have shown that the change in the relation between disease and functional status differs across age groups,^33–35 such that an increase in chronic diseases affects functioning more in younger age groups than in older age groups. Earlier, Freedman et al. showed that for older age groups many conditions became less debilitating and that the prevalence of disability among those with a particular condition declined.^36,37 Christensen et al. also concluded in their overview of challenges facing aging populations that the link between diseases and activity limitations or disability is weakening.⁸

Definitions of disability might explain why our findings regarding the trend in associations with diseases were different from the findings of others.³⁸ Disability is an umbrella term for different dimensions of functioning. The International Classification of Functioning, Disability, and Health model of the World Health Organization distinguishes body functions and structures, activities, and participation,³⁹ and the disablement process described by Verbrugge and Jette comprises pathology, impairments, functional limitations, and disability.⁴⁰ The use of aids and assistive devices does not alter activity limitations or impairments, but it does change the way these limitations affect people in performing their activities and in participating. For example, people with osteoarthritis have difficulties bending their knees (problem in body function and structure) and walking (activity limitation), but because they use a walking stick or other device, they still move around and do their shopping (participation). The extent to which an increasing prevalence in diseases is reflected in an increasing prevalence in limitations depends on which dimension of disability is assessed. It might well be that the improvements in health care and environmental factors affect participation but not activity limitations. The instruments we used in our study assessed disability mainly in the sense of activity limitations. This might explain why we did not find a clear change in the disabling impact of diseases. Future studies that distinguish between these dimensions in disability are needed to shed further light on this issue.

Another explanation for our finding that diseases became only slightly less disabling might be that the more objective functioning of the older population actually did improve, but we were not able to measure this change via our self-reports. Disability is a socially defined construct that depends not only on a person's activity limitations and the demands of the environment but also on that person's expectations of daily life.¹³ A change in these expectations over recent decades—perhaps because older persons are less inclined to accept deterioration in their functioning³³—might have concealed a true decrease in activity limitations. Measuring functional status more objectively with performance tests should contribute to our understanding of the changes in the relationships among diseases, objective functioning, and self-rated disability.

Methodological Issues

The major strength of our study is that we combined a number of large-scale surveys in the Netherlands, allowing us to study data collected over a long period by repeated cross-sectional and longitudinal surveys using the same method. Moreover, questionnaire items within surveys changed little over time, making trend analysis possible. Because of the heterogeneity among surveys, we thought the best approach was to use meta-analyses to combine the results from the different surveys to calculate best estimates of the time trends.

The Netherlands is a small country with a relatively homogeneous population, so we think it is unlikely that differences in coverage of regions affected the results.

Our study lacks information on institutionalized populations. The proportion of older people aged 55 to 85 years living in institutions decreased from 2.3% in 1995 to 1.2% in 2008.⁴¹ If this decrease is attributable to higher thresholds for institutionalizing people, that would imply that more people with diseases and activity limitations can be found in the community, meaning that our figures would mask a decline in disability prevalence. However, such effects, if they exist, are expected to be relatively small. We do not know the extent to which the exclusion of institutionalized populations might have affected our results.

There were large differences in the proportion of missing values among the surveys. Although we used multiple imputation to minimize any bias introduced by these missing values, the imputations have made the estimates less precise. This might explain why the decreasing time trends in at least moderate activity limitations and in the association between having 1 or more diseases and limitations were not statistically significant in MAAS.

The overall results in the trend for activity limitations differed for the 2 instruments: OECD limitations were stable, whereas SF-36 limitations decreased. Systematic differences between the OECD long-term disability indicator and the SF-36 might explain these differences.^7,23,24 Whereas OECD items ask about difficulty in the ability to perform actions, the SF-36 asks about limitations in executing activities because of impaired health. However, this difference cannot wholly explain the difference in trends, because one of the surveys that used OECD (POLS) also found the prevalence of activity limitations to be declining.

Our data suggest an overall increase in the prevalence of diseases since 1990 and no change or a slight decrease in activity limitations over the same time period. One explanation for these findings is that diseases have become less disabling; but, according to our results, this is only a small part of the explanation. It is also possible that the currently available data did not allow us to fully elucidate the changing relationship between diseases and their disabling consequences. Further research that uses performance tests to measure disability and that measures disability at the level of participation might shed more light on these trends.