Co-occurring medical conditions in adults with Down syndrome: A systematic review toward the development of health care guidelines

2.1 Survey of resources on health conditions

Using the National Library of Medicine (NLM) PubMed database (NCBI, 1946–2015), we undertook a survey of review articles that discussed the co-occurrence of medical conditions in adults with DS. Many of the 18 articles contained recommendations for routine screening or evaluation, but only a portion contained original, clinical data in support of their recommendations (Henderson et al., 2007; Jensen et al., 2013; Jones, 2009; Maatta et al., 2011; Real de Asua et al., 2015; van Allen et al., 1999; Van Buggenhout et al., 1999). Additional sources of reference not indexed in PubMed included books and book chapters (Chicoine & McGuire, 2010; Pueschel, 2006; Pueschel & Pueschel, 1992; Rubin & Dwyer, 1989) guidance documents prepared for health providers (Cohen & Group, 1999; Sullivan et al., 2006; Van Cleve et al., 2006), several journal articles (Kerins, Petrovic, Bruder, & Gruman, 2008; Prasher, 1994), and websites (Forster-Gibson & Berg, 2011). See Table 1a for a summary of the resources about general health that were consulted.

2.2 Survey of review articles

The 18 review articles identified through PubMed were reviewed in detail to determine the types of medical condition that were considered and discussed. Review articles were classified according to study design and the source of patient data as either “literature review/expert opinion,” “clinic chart review/original data,” “cohort survey/original data.” See Table 1b for a summary of data classification.

The frequency at which specific conditions were discussed was totaled across all the articles, and thereby served as a “post hoc consensus” of informed expert-opinion regarding the clinical significance of these conditions in adults with DS. This information was then used to inform the next stage of our review. See Table 2 for the frequency and rank order of medical conditions discussed in the review articles.

The conditions discussed in the publications in rank-order included ophthalmologic-vision, age-related dementia, behavior-mental health, thyroid disease, otolaryngology-hearing, cardiac disease, musculoskeletal-cervical spine, overweight-obesity, respiratory-sleep apnea, dermatologic concerns, seizures, dental concerns, gastrointestinal disorders, vaccination-infectious disease, gynecology-women's health, autoimmune disorders, type II diabetes and various cancers. Fewer, than one-third of articles discussed urologic-renal disorders, hematologic conditions, movement disorders, medication use, hyperlipidemia, gout, chronic pain, and syncope. None of the review articles discussed hospitalizations, end of life care, or cause of death.

2.3 Topic selection

Our Workgroup agreed that those conditions appearing in greater than 50% of the review articles warranted priority for further review. We prioritized those conditions that met criteria outlined by the United States Preventive Service Task-Force (USPSTF) because of (1) public health importance (i.e., burden of suffering and expected effectiveness of the preventive service to reduce that burden) and (2) the potential for recommendations to impact clinical practice (based on existing controversy or the belief that a gap exists between evidence and practice) (USPSTF, 2008). It was the consensus of the Workgroup to proceed with review of the following medical topics initially, thyroid disease, hearing impairment, congenital heart disease, cervical spine disease, osteopenia-osteoporosis, overweight-obesity, and sleep apnea. The Workgroup continues to evaluate the literature on visual impairment, behavior-mental health, age-related dementia, pulmonary disease, dermatology, gastrointestinal problems, dental problems, infectious disease, and women's health to be included in future manuscripts.

2.4 Key questions

In accordance with USPSTF practice, we next formulated a series of key questions. The formulation of key questions is an integral part of the approach to conducting systematic literature reviews. Along with the analytic framework, these questions specify the logic and scope of the topic and become critical in guiding the literature search, abstraction, and analysis process (USPSTF, 2008). By consensus, the Workgroup agreed that key questions needed to pertain to the clinical prevalence, severity, risk-factors, screening or evaluation methods, and potential benefits and/or harms in an adult population of persons with DS.

2.5 PubMed literature search

A second phase of topical literature searches were conducted in 2015–2016 also using the National Library of Medicine (NLM) biomedical literature database PubMed (MEDLINE) (NCBI, 1946–2013) to identify original research manuscripts addressing our prioritized topics. We used the Medical Subject Headings (MeSH) (the NLM controlled vocabulary thesaurus for indexing) to capture related entry terminology in our searches. For example, the MeSH term “Down syndrome” included the search entry terms: Downs syndrome, Down's syndrome, Mongolism, Trisomy 21, Partial Trisomy 21.

The MESH term “Down syndrome” was combined with one or more MeSH main heading terms to capture literature (unfiltered) about specific conditions in our search. Then, the limiters “Human,” “>19 years” were applied to narrow the scope of the search (filtered). Abstracts from Medline were reviewed and excluded according to their relevance in pertaining to key questions. Whenever an abstract made mention of any key question (or there was doubt) the full article was procured. The sections 2 and 3 were then reviewed to determine which articles met inclusion or exclusion criteria. A single reviewer from our group was chosen to conduct the literature searches then individual reviewers performed the data review and extraction process. See Table 3 for results of PubMed searches.

2.6 Article inclusion criteria

Study sample includes those >19 years, data addresses at minimum one key question, supporting data are original (not previously published), case series includes >5 participants, or uses a cohort, case-series or case-control research design or randomized clinical trial.

2.7 Exclusion criteria

Study sample includes those <18 years (exclusively), data do not address at least one key question, study uses an uninterpretable methodology, data have been previously published or does not provide supporting data.

2.8 Data extraction by condition

Using only the PubMed articles meeting inclusion, data pertaining to key questions were extracted from the Abstract section, sections 2 and 3, and entered into a preformatted Excel data template for analysis. See Table 4 for a summary of the articles used for the data extraction.

2.9 Evidence ratings by condition

Next a critical appraisal of each of the included articles was performed by reviewers to determine the type of research design used, subject ascertainment, total number of study subjects, source of control subjects, and the extent of internal validity and external validity. The grading of internal validity considers study design factors such as ascertainment and selection bias, test procedures and consideration of confounding variables; while external validity considers the generalizability of findings to a broader (more representative) population (USPSTF, 2008). See appendix VII in the USPSTF report for criteria on research design hierarchy, and the grading system used for scoring internal and external validity. See Table 5 for summary of evidence rating.

3.1 Thyroid disease

Of the nineteen articles reviewed, five used a case-control design (ll-2) (Hestnes et al., 1991; Kanavin, Aaseth, & Birketvedt, 2000; Kinnell, Gibbs, Teale, & Smith, 1987; Murdoch, Ratcliffe, McLarty, Rodger, & Ratcliffe, 1977; Percy et al., 1990) while the remaining fourteen were cohort studies (ll-2) (Baxter et al., 1975; Dinani & Carpenter, 1990; Kohen & Wise, 1992; Korsager, Chatham, & Ostergaard Kristensen, 1978; Percy et al., 2003; Prasher & Haque, 2005; Van Buggenhout et al., 1999) or case series (lll) (Feingold, 2004; Jensen et al., 2013; Mani, 1988; Percy et al., 2003; Prasher, Ninan, & Haque, 2011; Real de Asua et al., 2015; Tenenbaum et al., 2012) focused exclusively on individuals with DS. From the case-control studies, one study used controls with psychiatric disease from a residential facility (Murdoch et al., 1977), two studies used controls with other intellectual disabilities (ID) (Hestnes et al., 1991; Kanavin et al., 2000), and two studies used typically developing controls (Kinnell et al., 1987; Percy et al., 1990). The cumulative number of DS subjects studied appears sufficient (N = 1426) having been ascertained from residential institutions (44%), community samples (45%) and clinics or unspecified sources (11%). Eleven of the articles were published prior to the year 2000 (Table 4).

The scope of evaluation included standard thyroid function tests and/or anti-thyroid antibody titers. The medical comorbidities assessed in the studies included treatment with thyroxine (Baxter et al., 1975; Feingold, 2004; Mani, 1988; Prasher et al., 2011), presence of anti-thyroid antibodies, other autoimmune conditions (Real de Asua et al., 2015; Tenenbaum et al., 2012), or dementia (Percy et al., 1990; Tenenbaum et al., 2012; Van Buggenhout et al., 1999).

The prevalence of thyroid disease, including both hypothyroidism and subclinical hypothyroidism, appears to be higher in adults with DS (27% across studies) compared to those in the general population. There is only limited evidence regarding the prevalence of hyperthyroidism (estimated 3% across studies) (Hestnes et al., 1991; Kinnell et al., 1987; Percy et al., 1990; Real de Asua et al., 2015; Van Buggenhout et al., 1999). Data on severity of thyroid disease in adults with DS are limited, but the case-control studies suggest significant differences in thyroid function test values compared to controls. Overall, a high burden of thyroid disease is evident in this population, and is further supported by the high prevalence of thyroid disease in children with DS (Roizen et al., 2014).

Although conditions such as autoimmune disease are common in people with DS, there is a lack of studies exploring the relationship of these conditions with thyroid disease. One study noted a prevalence of thyroid disease in 74% of a sample of 136 children with diabetes and DS (Aitken et al., 2013), but studies about the co-occurrence of thyroid disease and other autoimmune conditions in adults are limited (Prasher, 1999).

We assigned internal validity ratings of good to 15 and fair to 4 of the studies based upon design considerations. Seven studies received external validity ratings of good, four were fair while eight were rated as poor. External validity was limited in several of the studies whenever participants were recruited from institutional settings which increased the likelihood for more serious comorbid medical conditions (Table 5).

3.2 Cervical spine

Of the articles reviewed, 15 addressed atlanto-axial instability (AAI) and 5 addressed degenerative disease of the cervical spine. The total number of adults with DS studied was large (N = 1,561), but only three studies utilized controls. Thirteen of the studies used a cohort or case series design (lll), while three used a case-control design (ll-2) (Alvarez & Rubin, 1986; Burke, French, & Roberts, 1985; Cooke & Lansdall-Welfare, 1991; El-Khouri et al., 2014; Elliott, Morton, & Whitelaw, 1988; Ferguson et al., 1997; French, Burke, Roberts, Whitecloud, & Edmunds, 1987; MacLachlan et al., 1993; Miller, Capusten, & Lampard, 1986; Miller, Grace, & Lampard, 1986; Morton, Khan, Murray-Leslie, & Elliott, 1995; Pueschel et al., 1987; Pueschel, Scola, & Pezzullo, 1992; Roy, Baxter, & Roy, 1990; Tangerud, Hestnes, Sand, & Sunndalsfoll, 1990; Van Dyke & Gahagan, 1988) (Table 4).

The scope of evaluation included measurement of the atlanto-dens interval (ADI) or bone height taken from plain films without consideration of co-morbid medical conditions. Although there was some variation in the measures used to define increased ADI, most studies used distances between 4.5 and 5 mm.

The prevalence of AAI in adults with DS (2–20%) appears to be decreased compared to children with DS (15–20%) but higher than typical age-matched controls (Alvarez & Rubin, 1986; Burke et al., 1985; Cooke & Lansdall-Welfare, 1991; El-Khouri et al., 2014; Elliott et al., 1988; Ferguson et al., 1997; French et al., 1987; Miller, Capusten, et al., 1986; Pueschel et al., 1987; Roy et al., 1990; Tangerud et al., 1990). The article with the highest prevalence of AAI (20%) used a cut off of 4 mm which may in part explain the findings (Miller, Capusten, et al., 1986). The presence of os odontoideum and/or ossicles appears to be a marker of high-risk in adults as it is in children (Burke et al., 1985; El-Khouri et al., 2014). Males and females appear to have similar risk; however, periods of inflammation may increase risk (Pueschel et al., 1987).

The prevalence of spondylosis or degenerative change of the cervical spine appears to be increased (33–64%) among adults with DS compared to the general population (Burke et al., 1985; MacLachlan et al., 1993; Miller, Capusten, et al., 1986; Tangerud et al., 1990; Van Dyke & Gahagan, 1988). These changes often localize to higher levels of the cervical spine and appear to increase in severity with age (Miller, Capusten, et al., 1986; Tangerud et al., 1990).

Three studies received internal validity ratings of fair and thirteen were rated as good based on design considerations. Ten studies received external validity ratings of fair, while six received a good rating. These assignments reflect the frequent ascertainment of samples from institutionalized settings (Table 5).

3.3 Hearing impairment

Four of the articles on hearing loss focused exclusively on persons with DS. Five articles also used non-DS controls with intellectual disability (ID), and one included healthy individuals from the general population. Six articles were cohort studies (II-2) (Lavis, 1997; Evenhuis, van Zanten, Brocaar, & Roerdinkholder, 1992; Keiser, Montague, Wold, Maune, & Pattison, 1981; Maatta et al., 2011; Van Buggenhout et al., 1999; van Schrojenstein Lantman-de Valk et al., 1994), three were case control (II-2) (Buchanan, 1990; Hassmann, Skotnicka, Midro, & Musiatowicz, 1998; Lowe & Temple, 2002), and one was a cross sectional study (lll) (Meuwese-Jongejeugd et al., 2006). The cumulative number of DS subjects studied appears large (N = 1,201). Seven of the articles were published prior to the year 2000 (Table 4).

The scope of evaluation entailed using standard audiologic methods such as pure tone audiometry or sound field testing (80%), however, various other methods were also used across studies suggesting variability in the approach to screening and evaluation. No consideration was given to other medical comorbidities.

The prevalence of any hearing impairment in adults with DS is up to 73% which is increased compared to both the general population and those with other forms of ID (Lavis, 1997; Meuwese-Jongejeugd et al., 2006). Disease severity appears to increase with age and up to 100% of DS adults experience hearing loss by 60 years, which indicates a high burden of disease in this population. Further support is evident from the high prevalence of middle ear disease and hearing impairment in children with DS (Roizen et al., 2014).

Eight studies received an internal validity rating of good and two were rated as fair. Good ratings on external validity was assigned to two studies while eight were rated as fair, based largely upon the consistently increased rates of hearing impairment in DS individuals when compared to those without DS across all studies (Table 5).

3.4 Overweight-obesity

Three of the studies on overweight-obesity were based on a case-control design (ll-2), two were case-series studies (ll-2). Two of the studies were limited only to persons with DS, while three employed contemporaneous non-DS controls with other forms of ID. All of the studies utilized large study samples (range 183–6,429). The cumulative number of subjects with DS studied was large (N = 1,426). Three of the articles were published prior to the year 2000. The scope of evaluation focused exclusively on measures of obesity itself, calculated body mass index (BMI) with no emphasis on comorbid medical conditions. Four studies received an internal validity rating of good, and one was rated as fair (Table 4).

Each of the studies utilized BMI as weight (kg)/height (m²) as the preferred method of evaluation for obesity (Bell & Bhate, 1992; Melville, Cooper, McGrother, Thorp, & Collacott, 2005; Prasher, 1995; Rubin, Rimmer, Chicoine, Braddock, & McGuire, 1998; Stancliffe et al., 2011). In four of the studies, participants were living in their home or community and recruited through a regional hospital or center-based medical clinic. A total of 412 males and 377 females with DS (total = 789), and 201 male and 171 female control subjects (total = 372) with other ID were studied. All subjects were between the ages of 15 and 76 years. Across these four studies 38% of DS subjects were classified as obese and 34% as overweight. Females were more likely than males (43% vs. 33%) to be obese, and about as likely to be overweight (32% vs. 35%). Thus 75% females and 68% males with DS were classified as overweight or obese. In the two studies that utilized ID control subjects (Bell & Bhate, 1992; Melville et al., 2005) 60% of females and 50% of males with ID were classified as overweight or obese. Additionally, in two of the studies a decline in BMI was noted with advancing age beyond 35 years (Prasher, 1995; Rubin et al., 1998).

In the one study that utilized an existing database of individuals with ID, that included persons with DS (N = 706), and those with non-specified ID (N = 5,627), 73% of both men and women with DS were classified as overweight or obese, compared to 65% of those with other ID (Stancliffe et al., 2011). Among those with DS women were more likely to be obese than men (48% vs. 41%); which was higher than women and men with non-specified ID (40% vs. 31%).

The external validity or generalizability of these findings to the larger population of adults with DS warrants a rating of good in four of the studies; and receives further support from the large number of community-residing persons who participated (Table 5). Additional support for generalizability stems from the consistently high prevalence of overweight-obesity (60–75%) across all studies, and the finding that obesity is often present by adolescence in youth with DS (Tenenbaum et al., 2011).

3.5 Congenital heart

Three of the studies on adult outcome of CHD were based on chart reviews of a DS cohort without controls (ll-2) (Majdalany et al., 2010; van Allen et al., 1999; Vis et al., 2010). The largest study, also retrospective used a case-control design and included a large numbers of participants with CHD both with/without DS (Baraona, Gurvitz, Landzberg, & Opotowsky, 2013). Information about financial costs, hospital LOS, non-cardiac comorbidities, and mortality was presented. A single study employed prospective cardiac screening in a subset of their participants ascertained retrospectively (Vis et al., 2010). Participants were ascertained through convenience samples including, medical clinics, hospitals and residential facilities. The cumulative number of subjects with DS reviewed was large (N = 10,472). All articles were published between 1999 and 2013 (Table 4).

The data included standard measures of cardiac function (echocardiography) or in one study morbidity (hospitalization, length of stay, medical conditions, and need for cardiac procedure) and mortality. Non-cardiac medical conditions were considered in only two studies.

Up to 17% of patients residing in a residential setting (the Netherlands) had undiagnosed CHD in addition to the 16% with previously CHD (33%). Regurgitation of the mitral, aortic, and tricuspid valves was present in 75% of subjects (Vis et al., 2010). In patients with AVSD repair left AV valve insufficiency and left ventricle outflow tract obstruction are the most frequently reported long-term complications requiring surgical repair (Martinez-Quintana, Rodriguez-Gonzalez, Medina-Gil, Agredo-Munoz, & Nieto-Lago, 2010). Hospitalized patients with DS and CHD had an increase prevalence of pulmonary hypertension, cyanosis and secondary erythrocytosis compared to those without the condition (Baraona et al., 2013). Among all hospitalized patients with CHD, mortality was higher for those with DS. Bacterial and aspiration pneumonia were exclusively associated with higher mortality in DS. Cardiac procedures, however, were less often performed in patients with DS.

Two studies received an internal validity rating of fair one was rated as good and another as poor. The external validity or generalizability of the findings received a rating of fair in three studies and one as poor (Table 5). Further support for these findings derives from the large number of participants surveyed, and the known prevalence (40–50%) of CHD in newborns with DS (Roizen et al., 2014).

3.6 Sleep apnea

Three of the studies of sleep apnea (OSA) focused on persons with DS exclusively, and just a single study employed non-DS (historical) controls also suspected of having OSA. One of the studies used a case-control design (II-2) (Trois et al., 2009), and two were small case series (III) of between 6 and 12 subjects (Andreou, Galanopoulou, & Gourgoulianis, 2002; Resta et al., 2003). Another utilized DS subjects, with/without comorbid depression as a within-syndrome case-control design (Capone, Aidikoff, Taylor, & Rykiel, 2013). The cumulative number of DS subjects studied is quite small (n = 71) having been ascertained primarily as clinic convenience samples. All of the articles were published after the year 2000 (Table 4).

The scope of evaluations focused on objective findings (apnea-hypopnea index [AHI], 02% saturation) taken from overnight polysomnogram, with some emphasis on co-occurring medical conditions, such as obesity, thyroid disease and depression. The apparent increased prevalence (85%) and high symptom severity (mean AHI = 25.9/hr) reported across these studies suggests a high disease burden in adults with DS.

Two studies were given an internal validity rating of poor and two fair. Because adult persons with DS constitute a special population in the United States, the external validity or generalizability of these findings must be made to a larger community-based population of adults with DS. Thus, two studies warrant an external validity rating of poor and two are rated as fair. These ratings reflect the small sample sizes and selection bias inherent to clinically ascertained samples (Table 5). However, additional support derives from the high prevalence of OSA in children with DS (Churchill, Kieckhefer, Landis, & Ward, 2012; Hoffmire, Magyar, Connolly, Fernandez, & van Wijngaarden, 2014).

3.7 Osteopenia-osteoporosis

Six of the articles on bone density utilized a case-control design focusing on adults with DS, and controls with ID (II-2) (Angelopoulou et al., 2000, 1999; Baptista, Varela, & Sardinha, 2005; Guijarro, Valero, Paule, Gonzalez-Macias, & Riancho, 2008; Sakadamis, Angelopoulou, Matziari, Papameletiou, & Souftas, 2002; Tyler et al., 2000). Two studies relied on a retrospective chart review (III) (Schrager, Kloss, & Ju, 2007; van Allen et al., 1999). The cumulative number of DS subjects studied was (N = 342). The articles were published between 1999 and 2008 (Table 4).

The prevalence of osteopenia-osteoporosis appears increased among adults with DS compared to adults in the general population and those with other forms of ID. It is unclear if there is a corresponding increase in bone fracture, but some studies suggest this. Risk factors for low bone-mass density (BMD) were identified in five of the studies and included immobility, inactivity, low calcium and vitamin D, low sunlight exposure, hypogonadism, and seizure disorders. Interestingly males and females appear to be equally affected. Most of the studies measured BMD on dual energy X-ray absorptiometry (DXA) scan as the preferred method of evaluation. The DXA method for screening is widely available and without need of modification in the DS population.

Three studies were retrospective medical clinic chart reviews and the remainder utilized community-based samples. Two studies received an internal validity rating of fair and six were rated as poor. The external validity or generalizability of the findings was rated as fair in two studies and poor in six of the studies (Table 5).

3.8 Evidence gaps

There is only limited published evidence available to answer our key questions for each of the co-occurring conditions under review in the adult DS population. Many older studies are descriptive, utilizing small convenience samples and either typical controls or those with IDD as a comparison group for DS. Rarely were persons with DS, without the target condition, used as controls for determining possible risk factors or associated co-morbidities within a larger DS sample. Indeed, there is scant information about risk factors or the direct physiologic impact of other medical co-morbidities on target disease prevalence or severity. No studies have addressed the financial costs or risks/benefits of screening in asymptomatic individuals. A single study addressed morbidity, mortality, and the financial costs associated with the specified medical condition (CHD). Due to the absence of longitudinal, prospective cohort data, we are unable to determine the natural history of disease progression in asymptomatic individuals who eventually become symptomatic, especially those presenting with severe disease. Studies about the use of standard versus alternative methodologies for screening were also generally unavailable.

3.9 Condition specific considerations and emerging knowledge

4.1 Limitations of the study

Limitations of this study include the restriction of our review only to that literature written in English and available through the NLM PubMed. Second, the number of studies available for review was limited and the quality of those studies was generally poor to fair, especially for those studies relying on data collected retrospectively from chart reviews, or when using convenience samples without adequate controls. Many studies were performed in a medical or residential setting because that is where one finds large numbers of adult individuals with DS. Thus, any resulting ascertainment bias may tend to oversample symptomatic individuals with severe disease. Perhaps this ascertainment bias reflects a common experience among adult primary care providers who are expected to manage patients with DS and complex symptomatology. It is these providers who will benefit most from having clinical guidance documents to assist in patient care and management. Third, a single reviewer extracted the data from each article and summarized the findings before it was re-reviewed by a panel of expert practitioners experienced in caring for adults with DS. Inter-rater reliability was not assessed. Despite these limitations, the study represents a coordinated effort by leading medical experts to critically review and synthesize the existing and emerging knowledge to best inform health screening and evaluation practices for adults with DS.

4.2 The adult population in perspective

The number of persons with DS living in the United States (2008–2010) is estimated to be between 200,000 and 250,000 (de Graaf et al., 2017; Presson et al., 2013); and the number of adults (>18 yr) with DS living in the United States approaches or exceeds 125,000 individuals.

As longevity continues to increase it is also expected that greater numbers of adults with DS will live to be of advanced-age (>45 yr) (Bittles & Glasson, 2004). This presents ongoing challenges to the primary care physicians expected to manage an array of congenital, chronic and age-related conditions (Bittles et al., 2007). Of the co-morbid health conditions typically mentioned, visual and hearing impairment, thyroid disease, obesity, sleep apnea, cardiopulmonary function, cervical and lumbar spondylosis, seizure disorder, and dementia are likely to remain the major considerations (Esbensen, 2010; Glasson et al., 2014). In this regard, guidance around end-of-life and palliative care also remains an area of need (McCallion et al., 2017).

4.3 Strategic planning

For planning purposes and informed by this review we estimate that the number of adults (>18 yr) with DS currently living in the United States with a specific co-occurring health condition can be determined by the following: estimated disease prevalence in the DS population (rounded up to the nearest 5%) × 125,000 estimated individuals (>18 yr) living in the United States = number of individuals with DS affected by the condition.

Thus, for hearing loss (100%) = 125,000; for sleep apnea (85%) = 106,250 individuals; for overweight-obesity (70%) = 87,500; for cervical spondylosis (60%) = 75,000; for osteopenia-osteoporosis (50%) = 62,500; for previously repaired or uncorrected CHD (35%) = 43,750; for thyroid disease (30%) = 37,500.

Due to the unique biologic underpinnings of trisomy 21 some medical conditions may exhibit unique features of etiology-pathogenesis and natural history compared to individuals without this chromosomal condition (Zigman, 2013). In clinical practice, multiple medical co-morbidities is the rule not the exception, and this requires difficult decision-making and management considerations (Evenhuis, Schoufour, & Echteld, 2013; Schoufour, Evenhuis, & Echteld, 2014). Taken together, these factors suggest a modified approach to both diagnosis and treatment in elderly or medically frail adults with DS. In such situations, assessment of the specific risks and potential benefits of diagnostic evaluation and its intended therapeutic purpose needs to be discussed openly with decision-makers. Management alternatives for those of advanced-age or nearing end-of-life need to be made available to healthcare providers and family decision makers to use as they see fit in their specific circumstances.

4.4 Toward guidelines

The single biggest challenge for guideline development is based upon their intended scope, breadth and depth. As DS is not a specific disease, but rather a unique human condition associated with a variety of developmental–anatomical differences, acquired (chronic) medical conditions, and precocious aging, such guidelines would potentially involve every major organ-system and life-stage experienced throughout adulthood. The best precedent for creating guidance documents has come from the efforts of the American Academy of Pediatrics (AAP, 2011). Although guidance beyond 21 yr is not within the scope of the AAP document, it never-the-less serves as an important educational tool about DS that would be of benefit to any physician. Much like an airline pilot flying without instrumentation, adult trained providers unfamiliar with DS or the AAP document remain in the unwelcome position of practicing medicine without any guidance whatsoever.

The intended audience for adult guidelines requires thorough consideration throughout the development process. Primary health care providers (physicians, nurses, nurse practitioners, and physician assistants) who will be providing direct care to adults are a primary target group (Qaseem, Snow, Owens, & Shekelle, 2010). The other stakeholder groups include caretakers (parents, siblings, and agency workers) and advocacy organizations (national and regional parent groups) who will use this information to advocate for quality health care locally and nationally (IOM, 2011). Stakeholders should be included in the review process particularly in determining whether an assessment of benefits, harms and potential alternative options are fully addressed (Diaz Del Campo et al., 2011). Deployment of invested stakeholders will be critical to the prompt dissemination and successful adoption of health guidelines in both the public health and primary care settings (Luke, Wald, Carothers, Bach, & Harris, 2013).

4.5 Realigning clinical research

Based simply on population prevalence in the United States DS is on the cusp of being considered a rare disease (frequency <200,000) (NIH, 2017). Further, it is likely that the prevalence rate for most co-occurring conditions is well within the range of rare disease designation. And so, it remains challenging to plan, organize, fund and enroll sufficient numbers of adult participants into existing data collection efforts and screening protocols, in part because of their numbers and geographical distribution.

It is not known what percent of the estimated 125,000 adults with DS living in the United States utilize services at an existing specialty clinic. Those who do almost certainly receive more comprehensive care compared to those who do not (Jensen et al., 2013; Skotko, Davidson, & Weintraub, 2013). Although the number of DS clinics serving the needs of adults are few, many are located at large, university-affiliated medical, research and training centers (AUCD, 2017; DSMIG-USA, 2017). Despite such advantages, clinical research on adults has not kept pace with the need for relevant information. What is required are better efforts to organize and support existing clinical DS programs to collect and share information on medical screening, diagnostic evaluation and treatment outcomes, as routinely performed at the point of care. Recent efforts to conduct multicenter data collection and sharing using clinician input data are successfully underway (Lavigne et al., 2017, 2015) and may provide the necessary mechanism for further progress if properly funded. Efforts to engage the larger community of families living with DS to participate in clinical research studies is also underway (Peprah et al., 2015). However, the availability of research funding commensurate with stated long-term goals has never materialized (NICHD, 2014).

Presently, the availability of dedicated research personnel and lack of infrastructure support each represent limiting factors in advancing a truly comprehensive data collection effort and person-centered research strategy. While the provision of high quality clinical care to persons with DS is challenging enough, it is yet another matter to capture this experience for the purpose of informing evidence-based care (Murillo, Reece, Snyderman, & Sung, 2006). With the necessary support and leadership it is well within the capacity of existing clinical programs to step up and address this urgent need (Carfi et al., 2015; McCabe, Hickey, & McCabe, 2011; Real de Asua et al., 2015).