*2.2. Statistical Analyses*

*Article* 

**and Herminia Diana Rosas 1,2,3**

Descriptive summaries were computed by age of AD diagnosis (early, typical, or late). Continuous variables were summarized using either the mean/standard deviation (SD) or using the median and interquartile range (IQR, 25th–75th percentile). If missing data were observed, the frequency of non-missing variable responses was augmented to the summaries of the continuous variables (e.g., mean (SD); n or median (IQR); n). Categorical variables were summarized as percentages and frequencies of non-missing responses. Using a complete-case analysis, preliminary differences in the distributions of categorical and continuous variables by diagnosis were assessed using either the chisquared test/Fisher's exact test or the Mann–Whitney test, respectively. The Spearman correlation coefficient and its 95% confidence interval (CI) were computed to summarize the monotonic relationship between the ages of hypothyroidism and AD onset. 6 College of Arts & Sciences, Boston University, Boston, MA 02215, USA; herschgi@bu.edu \* Correspondence**:** flai@partners.org **Abstract:** Adults with Down syndrome (DS) have an exceptionally high frequency of Alzheimer disease (AD) with a wide variability in onset, from 40 to 70 years of age. Equally prevalent in DS is hypothyroidism. In this study, we sought to quantify the relationship between the two. A total of 232 adults with DS and AD were stratified into three AD onset age groups: early (<47 years), typical (48–59), and late (>59). Among patients with available data, differences in the distributions of demographics, hypothyroidism variables (presence, age of onset), thyroid function tests, thyroid autoantibodies, and APOE genotypes were assessed (e.g., chi-squared, Mann–Whitney tests). Spearman and partial Spearman correlations and ordinal logistic regression models were constructed to **Citation:** Lai, F.; Mercaldo, N.D.; **Association between Hypothyroidism Onset and Alzheimer Disease Onset in Adults with Down Syndrome Florence Lai 1,2,3,\*, Nathaniel D. Mercaldo 1,2, Cassandra M. Wang 4, Micaela S. Hersch 5, Giovi G. Hersch 6**  1 Harvard Medical School, Boston, MA 02115, USA; nmercaldo@mgh.harvard.edu (N.D.M.); **Association between Hypothyroidism Onset and Alzheimer Disease Onset in Adults with Down Syndrome Florence Lai 1,2,3,\*, Nathaniel D. Mercaldo 1,2, Cassandra M. Wang 4, Micaela S. Hersch 5, Giovi G. Hersch 6**  1 Harvard Medical School, Boston, MA 02115, USA; nmercaldo@mgh.harvard.edu (N.D.M.); **Association between Hypothyroidism Onset and Alzheimer Disease Onset in Adults with Down Syndrome Florence Lai 1,2,3,\*, Nathaniel D. Mercaldo 1,2, Cassandra M. Wang 4, Micaela S. Hersch 5, Giovi G. Hersch 6 and Herminia Diana Rosas 1,2,3** 1 Harvard Medical School, Boston, MA 02115, USA; nmercaldo@mgh.harvard.edu (N.D.M.);

Partial Spearman correlation coefficients were also computed between ages of hypothyroidism onset and age of AD onset, while separately accounting for each demographic variable or co-varying medical conditions of interest (sex, APOE quantify the association between ages of AD and hypothyroidism onset with and without covariate adjustments. We observed a positive association between the ages of AD and hypothyroidism onset after accounting for APOE-Ɛ4 (correlation: 0.44, 0.24, 0.60; odds ratio: 1.09, 1.05–1.14). However, an early age of hypothyroidism onset and the presence of the APOE-Ɛ4 allele were independently associated with the early age of AD onset. Similar findings were observed when accounting for other factors. Our study provides evidence for the importance of hypothyroidism and associated pathological mechanisms for risk of AD in DS. **Keywords:** Down syndrome; early-onset Alzheimer disease; late-onset Alzheimer disease; hypothyroidism; thyroid autoantibodies; TSH; Free T4; APOE Ɛ4 Wang, C.M.; Hersch, M.S.; Hersch, G.G.; Rosas, H.D.; Association between Hypothyroidism Onset and Alzheimer Disease Onset in Adults with Down Syndrome. *Brain Sci.*  **2021**, *11*, x. https://doi.org/10.3390/xxxxx Academic Editor: Rebecca Sims Received: 31 July 2021 4 status, body mass index (BMI), history of vitamin B12 deficiency, history of obstructive sleep apnea (OSA), and level of intellectual disability). An exploratory series of proportional odds logistic regression models were constructed to quantify the association between the age of AD onset and age of hypothyroidism onset while separately accounting for each demographic or covarying medical condition and their interactions. Wald tests were performed to assess model complexity (non-linear terms, interaction effects). Parameter estimates, 95% CI, and *p*-values were computed to summarize the final regression models. rosas@helix.mgh.harvard.edu (H.D.R.) 2 Massachusetts General Hospital, Boston, MA 02114, USA 3 McLean Hospital, Belmont, MA 02478, USA 4 Harvard College, Harvard University, Cambridge, MA 02138, USA; cassandrawang@college.harvard.edu 5 School of Nursing, Simmons University, Boston, MA 02115, USA; micaela.hersch@simmons.edu 6 College of Arts & Sciences, Boston University, Boston, MA 02215, USA; herschgi@bu.edu \* Correspondence**:** flai@partners.org **Abstract:** Adults with Down syndrome (DS) have an exceptionally high frequency of Alzheimer disease (AD) with a wide variability in onset, from 40 to 70 years of age. Equally prevalent in DS is rosas@helix.mgh.harvard.edu (H.D.R.) 2 Massachusetts General Hospital, Boston, MA 02114, USA 3 McLean Hospital, Belmont, MA 02478, USA 4 Harvard College, Harvard University, Cambridge, MA 02138, USA; cassandrawang@college.harvard.edu 5 School of Nursing, Simmons University, Boston, MA 02115, USA; micaela.hersch@simmons.edu 6 College of Arts & Sciences, Boston University, Boston, MA 02215, USA; herschgi@bu.edu \* Correspondence**:** flai@partners.org **Abstract:** Adults with Down syndrome (DS) have an exceptionally high frequency of Alzheimer disease (AD) with a wide variability in onset, from 40 to 70 years of age. Equally prevalent in DS is rosas@helix.mgh.harvard.edu (H.D.R.) 2 Massachusetts General Hospital, Boston, MA 02114, USA 3 McLean Hospital, Belmont, MA 02478, USA 4 Harvard College, Harvard University, Cambridge, MA 02138, USA; cassandrawang@college.harvard.edu 5 School of Nursing, Simmons University, Boston, MA 02115, USA; micaela.hersch@simmons.edu 6 College of Arts & Sciences, Boston University, Boston, MA 02215, USA; herschgi@bu.edu \* Correspondence**:** flai@partners.org **Abstract:** Adults with Down syndrome (DS) have an exceptionally high frequency of Alzheimer disease (AD) with a wide variability in onset, from 40 to 70 years of age. Equally prevalent in DS is

### Accepted: 8 September 2021 **3. Results** hypothyroidism. In this study, we sought to quantify the relationship between the two. A total of 232 adults with DS and AD were stratified into three AD onset age groups: early (<47 years), typical hypothyroidism. In this study, we sought to quantify the relationship between the two. A total of 232 adults with DS and AD were stratified into three AD onset age groups: early (<47 years), typical hypothyroidism. In this study, we sought to quantify the relationship between the two. A total of 232 adults with DS and AD were stratified into three AD onset age groups: early (<47 years), typical

### Published: 16 September 2021 *3.1. Demographics* (48–59), and late (>59). Among patients with available data, differences in the distributions of de-(48–59), and late (>59). Among patients with available data, differences in the distributions of de-(48–59), and late (>59). Among patients with available data, differences in the distributions of de-

**1. Introduction**  Adults with Down syndrome (DS) have an especially high risk for developing Alzheimer disease (AD), with onset at least two decades earlier than in the general population [1,2]. Although there is great variation in the age of AD onset, from as early as 40 to as late as 70 or older [3,4], this variability in onset is poorly understood. The leading hypothesis for the pathogenesis of AD in DS has been attributed to overexpression of the gene for amyloid precursor protein (APP) located on the triplicated chromosome 21 [5], alt-**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Demographics for the cohort and for each diagnostic group are provided in Table 1. The cohort included 36 individuals with early AD onset, 160 with a typical age of AD onset and 36 with late AD onset. Differences in the distribution of sex and the level of intellectual disability were not observed across the three age of AD onset groups (*p* = 0.776 and *p* = 0.265, respectively). There was a significantly higher frequency of patients carrying an APOE mographics, hypothyroidism variables (presence, age of onset), thyroid function tests, thyroid autoantibodies, and APOE genotypes were assessed (e.g., chi-squared, Mann–Whitney tests). Spearman and partial Spearman correlations and ordinal logistic regression models were constructed to quantify the association between ages of AD and hypothyroidism onset with and without covariate adjustments. We observed a positive association between the ages of AD and hypothyroidism onset after accounting for APOE-Ɛ4 (correlation: 0.44, 0.24, 0.60; odds ratio: 1.09, 1.05–1.14). However, an early age of hypothyroidism onset and the presence of the APOE-Ɛ4 allele were independently as-4 allele ( mographics, hypothyroidism variables (presence, age of onset), thyroid function tests, thyroid autoantibodies, and APOE genotypes were assessed (e.g., chi-squared, Mann–Whitney tests). Spearman and partial Spearman correlations and ordinal logistic regression models were constructed to quantify the association between ages of AD and hypothyroidism onset with and without covariate adjustments. We observed a positive association between the ages of AD and hypothyroidism onset after accounting for APOE-Ɛ4 (correlation: 0.44, 0.24, 0.60; odds ratio: 1.09, 1.05–1.14). However, an early age of hypothyroidism onset and the presence of the APOE-Ɛ4 allele were independently as-3/4 and mographics, hypothyroidism variables (presence, age of onset), thyroid function tests, thyroid autoantibodies, and APOE genotypes were assessed (e.g., chi-squared, Mann–Whitney tests). Spearman and partial Spearman correlations and ordinal logistic regression models were constructed to quantify the association between ages of AD and hypothyroidism onset with and without covariate adjustments. We observed a positive association between the ages of AD and hypothyroidism onset after accounting for APOE-Ɛ4 (correlation: 0.44, 0.24, 0.60; odds ratio: 1.09, 1.05–1.14). However, an early age of hypothyroidism onset and the presence of the APOE-Ɛ4 allele were independently as-4/4) in the early AD onset group than in the later AD onset groups (*p* = 0.040).

hough factors other than amyloid likely influence the wide range of AD onset age in those with DS, just as they do in sporadic AD. Some of these factors include not only the Apolipoprotein Ɛ4 (APOE Ɛ4) genotype, but other genetic factors, as well as environmental or biological factors and co-existing medical conditions. Thyroid dysfunction, including congenital, subclinical, and autoimmune thyroid **Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creasociated with the early age of AD onset. Similar findings were observed when accounting for other factors. Our study provides evidence for the importance of hypothyroidism and associated pathological mechanisms for risk of AD in DS. **Keywords:** Down syndrome; early-onset Alzheimer disease; late-onset Alzheimer disease; hyposociated with the early age of AD onset. Similar findings were observed when accounting for other factors. Our study provides evidence for the importance of hypothyroidism and associated pathological mechanisms for risk of AD in DS. **Keywords:** Down syndrome; early-onset Alzheimer disease; late-onset Alzheimer disease; hyposociated with the early age of AD onset. Similar findings were observed when accounting for other factors. Our study provides evidence for the importance of hypothyroidism and associated pathological mechanisms for risk of AD in DS. **Keywords:** Down syndrome; early-onset Alzheimer disease; late-onset Alzheimer disease; hypo-A history of hypothyroidism was present in 58.3% of the early, 65.6% of the typical and 72.2% of the late AD onset cases; there was no significant difference in the frequency of hypothyroidism amongst groups (*p* = 0.463). There was no significant difference in BMI at the time of AD diagnosis (*p* = 0.970). There was, however, a higher prevalence of a diagnosis of OSA in the early AD onset group (*p* = 0.048).

*Brain Sci.* **2021**, *11*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/brainsci conditions [6], is a very common medical co-morbidity in individuals with DS. In a large tivecommons.org/licenses/by/4.0/). thyroidism; thyroid autoantibodies; TSH; Free T4; APOE Ɛ4 **1. Introduction**  Adults with Down syndrome (DS) have an especially high risk for developing Alzheimer disease (AD), with onset at least two decades earlier than in the general population [1,2]. Although there is great variation in the age of AD onset, from as early as 40 to as late as 70 or older [3,4], this variability in onset is poorly understood. The leading hypothesis for the pathogenesis of AD in DS has been attributed to overexpression of the gene thyroidism; thyroid autoantibodies; TSH; Free T4; APOE Ɛ4 **1. Introduction**  Adults with Down syndrome (DS) have an especially high risk for developing Alzheimer disease (AD), with onset at least two decades earlier than in the general population [1,2]. Although there is great variation in the age of AD onset, from as early as 40 to as late as 70 or older [3,4], this variability in onset is poorly understood. The leading hypothesis for the pathogenesis of AD in DS has been attributed to overexpression of the gene thyroidism; thyroid autoantibodies; TSH; Free T4; APOE Ɛ4 **1. Introduction**  Adults with Down syndrome (DS) have an especially high risk for developing Alzheimer disease (AD), with onset at least two decades earlier than in the general population [1,2]. Although there is great variation in the age of AD onset, from as early as 40 to as late as 70 or older [3,4], this variability in onset is poorly understood. The leading hypothesis for the pathogenesis of AD in DS has been attributed to overexpression of the gene **Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institu-Given previous reports of the known relationship between thyroid dysfunction and cognitive impairment [16,29], we sought to confirm that individuals were euthyroid at the time of AD diagnosis (Table 2). Among patients with a measured TSH value, we were unable to detect differences in the distribution of these values by age of AD onset (continuous: *p* = 0.616, categorized TSH: *p* = 0.310). Approximately 3% of the early, 7% of the typical, and 9% of the late AD onset groups had TSH levels less than 0.34 uIU/mL, suggesting a possible hyperthyroid state. Approximately 3% of the early, 15% of the typical, and 15% of the late AD onset patients had TSH levels greater than 5 ulU/mL, suggesting that they may have been inadequately treated at the time of AD diagnosis. Differences in the distributions of Free T4 levels were not detected amongst the three groups (*p* = 0.277).

> for amyloid precursor protein (APP) located on the triplicated chromosome 21 [5], although factors other than amyloid likely influence the wide range of AD onset age in those with DS, just as they do in sporadic AD. Some of these factors include not only the Apolipoprotein Ɛ4 (APOE Ɛ4) genotype, but other genetic factors, as well as environmen-

> conditions [6], is a very common medical co-morbidity in individuals with DS. In a large

*Brain Sci.* **2021**, *11*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/brainsci

tal or biological factors and co-existing medical conditions.

*Brain Sci.* **2021**, *11*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/brainsci

tal or biological factors and co-existing medical conditions.

for amyloid precursor protein (APP) located on the triplicated chromosome 21 [5], although factors other than amyloid likely influence the wide range of AD onset age in those with DS, just as they do in sporadic AD. Some of these factors include not only the Apolipoprotein Ɛ4 (APOE Ɛ4) genotype, but other genetic factors, as well as environmen-

for amyloid precursor protein (APP) located on the triplicated chromosome 21 [5], although factors other than amyloid likely influence the wide range of AD onset age in those with DS, just as they do in sporadic AD. Some of these factors include not only the Apolipoprotein Ɛ4 (APOE Ɛ4) genotype, but other genetic factors, as well as environmen-

conditions [6], is a very common medical co-morbidity in individuals with DS. In a large

*Brain Sci.* **2021**, *11*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/brainsci

tal or biological factors and co-existing medical conditions.

2 Massachusetts General Hospital, Boston, MA 02114, USA

1 Harvard Medical School, Boston, MA 02115, USA; nmercaldo@mgh.harvard.edu (N.D.M.);

5 School of Nursing, Simmons University, Boston, MA 02115, USA; micaela.hersch@simmons.edu

esis for the pathogenesis of AD in DS has been attributed to overexpression of the gene

**Association between Hypothyroidism Onset and Alzheimer** 

**Florence Lai 1,2,3,\*, Nathaniel D. Mercaldo 1,2, Cassandra M. Wang 4, Micaela S. Hersch 5, Giovi G. Hersch 6** 

**Disease Onset in Adults with Down Syndrome** 

rosas@helix.mgh.harvard.edu (H.D.R.)

cassandrawang@college.harvard.edu

\* Correspondence**:** flai@partners.org

*Article* 

**and Herminia Diana Rosas 1,2,3**

**Citation:** Lai, F.; Mercaldo, N.D.; Wang, C.M.; Hersch, M.S.; Hersch, G.G.; Rosas, H.D.; Association between Hypothyroidism Onset and Alzheimer Disease Onset in Adults with Down Syndrome. *Brain Sci.* 

https://doi.org/10.3390/xxxxx

Academic Editor: Rebecca Sims

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institu-

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Received: 31 July 2021 Accepted: 8 September 2021 Published: 16 September 2021

tional affiliations.

**2021**, *11*, x.


**Table 1.** Demographic summaries by age of AD onset group. 6 College of Arts & Sciences, Boston University, Boston, MA 02215, USA; herschgi@bu.edu

for amyloid precursor protein (APP) located on the triplicated chromosome 21 [5], alt-**Table 2.** Thyroid function tests: summaries by age of AD onset group.

