# Alzbiomarker

This interactive resource organizes decades of data on biomarkers for Alzheimer's disease. Biomarker measurements from the cerebrospinal fluid and blood are curated from the primary literature and meta-analyzed. The database was developed in collaboration with Henrik Zetterberg and colleagues at the University of Gothenburg and the meta-analysis results of the Version 1.0 dataset have been published (Olsson et al., 2016). We welcome your feedback and invite you to contact us at alzbiomarker@alzforum.org.

See below for links to each meta-analysis along with key statistics, such as p value and effect size -- a ratio of the mean biomarker concentration of one cohort over another (e.g. AD/control). View this page on a desktop, laptop, or tablet for a colorful, interactive graph of the results, plotted by effect size and z score.

Version 2.0, April 2017

275

Papers

1520

Cohorts

25

Biomarkers

36

Meta-Analyses

# Meta-Analyses

The graph, and the table below it, show key results from all meta-analyses. As the overall result, the effect size—a ratio of the mean biomarker concentration of one condition over another (e.g. AD/control)—is plotted against the z score, a statistic indicating how far away the observed ratio is from the null hypothesis of no difference. For example, a z score of 1.96 means that the observed ratio is 1.96 times larger than the standard error; it is equivalent to p=0.05. A z score of 3.89 is equivalent to p=0.0001. Customize the plot by selecting group comparisons and/or biomarkers. The size of each circle reflects the total number of subjects in the meta-analysis. Mouse over the circle to learn more. In the table, click the arrowheads to sort each data column.

## Biomarkers:

Meta-Analysis | Effect Size | z score | p value | Number of Comparisons |
Number of Subjects |
---|---|---|---|---|

AD vs CTRL: Aβ38 (CSF) | Effect Size: 0.955 | z score | p value: z = 1.05 | p = 0.29345 | Number of Comparisons: 11 | Number of Subjects: AD N=478, CTRL N=588 |

AD vs CTRL: Aβ40 (CSF) | Effect Size: 0.928 | z score | p value: z = 3.45 | p = 0.00056 | Number of Comparisons: 32 | Number of Subjects: AD N=1514, CTRL N=1285 |

AD vs CTRL: Aβ40 (Plasma and Serum) | Effect Size: 1.065 | z score | p value: z = 2.20 | p = 0.02803 | Number of Comparisons: 26 | Number of Subjects: AD N=2125, CTRL N=4265 |

AD vs CTRL: Aβ42 (CSF) | Effect Size: 0.559 | z score | p value: z = 49.55 | p <0.0001 | Number of Comparisons: 168 | Number of Subjects: AD N=11277, CTRL N=8315 |

AD vs CTRL: Aβ42 (Plasma and Serum) | Effect Size: 1.031 | z score | p value: z = 0.86 | p = 0.38718 | Number of Comparisons: 27 | Number of Subjects: AD N=2336, CTRL N=4452 |

AD vs CTRL: GFAP (CSF) | Effect Size: 1.119 | z score | p value: z = 0.34 | p = 0.73555 | Number of Comparisons: 2 | Number of Subjects: AD N=59, CTRL N=39 |

AD vs CTRL: MCP-1 (CSF) | Effect Size: 1.123 | z score | p value: z = 4.70 | p <0.0001 | Number of Comparisons: 5 | Number of Subjects: AD N=110, CTRL N=114 |

AD vs CTRL: MCP-1 (Plasma and Serum) | Effect Size: 0.999 | z score | p value: z = 0.02 | p = 0.98639 | Number of Comparisons: 6 | Number of Subjects: AD N=540, CTRL N=344 |

AD vs CTRL: NFL (CSF) | Effect Size: 2.313 | z score | p value: z = 8.92 | p <0.0001 | Number of Comparisons: 10 | Number of Subjects: AD N=271, CTRL N=322 |

AD vs CTRL: NSE (CSF) | Effect Size: 1.470 | z score | p value: z = 2.47 | p = 0.01358 | Number of Comparisons: 7 | Number of Subjects: AD N=258, CTRL N=160 |

AD vs CTRL: NSE (Plasma and Serum) | Effect Size: 0.999 | z score | p value: z = 0.01 | p = 0.99192 | Number of Comparisons: 3 | Number of Subjects: AD N=102, CTRL N=97 |

AD vs CTRL: VLP-1 (CSF) | Effect Size: 1.355 | z score | p value: z = 9.69 | p <0.0001 | Number of Comparisons: 8 | Number of Subjects: AD N=468, CTRL N=812 |

AD vs CTRL: YKL-40 (CSF) | Effect Size: 1.291 | z score | p value: z = 16.27 | p <0.0001 | Number of Comparisons: 14 | Number of Subjects: AD N=668, CTRL N=630 |

AD vs CTRL: YKL-40 (Plasma) | Effect Size: 1.950 | z score | p value: z = 1.93 | p = 0.05309 | Number of Comparisons: 3 | Number of Subjects: AD N=155, CTRL N=233 |

AD vs CTRL: albumin ratio (CSF/Serum and/or Plasma) | Effect Size: 1.103 | z score | p value: z = 2.23 | p = 0.02564 | Number of Comparisons: 21 | Number of Subjects: AD N=886, CTRL N=466 |

AD vs CTRL: hFABP (CSF) | Effect Size: 1.394 | z score | p value: z = 5.63 | p <0.0001 | Number of Comparisons: 5 | Number of Subjects: AD N=285, CTRL N=297 |

AD vs CTRL: hFABP (Plasma and Serum) | Effect Size: 1.049 | z score | p value: z = 0.40 | p = 0.69241 | Number of Comparisons: 2 | Number of Subjects: AD N=55, CTRL N=74 |

AD vs CTRL: neurogranin (CSF) | Effect Size: 1.937 | z score | p value: z = 6.71 | p <0.0001 | Number of Comparisons: 10 | Number of Subjects: AD N=574, CTRL N=555 |

AD vs CTRL: neurogranin (Plasma) | Effect Size: 0.714 | z score | p value: z = 1.14 | p = 0.25272 | Number of Comparisons: 2 | Number of Subjects: AD N=45, CTRL N=49 |

AD vs CTRL: sAPPα (CSF) | Effect Size: 1.026 | z score | p value: z = 0.59 | p = 0.55842 | Number of Comparisons: 11 | Number of Subjects: AD N=613, CTRL N=457 |

AD vs CTRL: sAPPβ (CSF) | Effect Size: 1.020 | z score | p value: z = 0.61 | p = 0.54244 | Number of Comparisons: 12 | Number of Subjects: AD N=672, CTRL N=481 |

AD vs CTRL: sTREM2 (CSF) | Effect Size: 1.272 | z score | p value: z = 3.92 | p <0.0001 | Number of Comparisons: 6 | Number of Subjects: AD N=388, CTRL N=370 |

AD vs CTRL: tau-phospho (CSF) | Effect Size: 1.882 | z score | p value: z = 29.47 | p <0.0001 | Number of Comparisons: 116 | Number of Subjects: AD N=8416, CTRL N=5957 |

AD vs CTRL: tau-total (CSF) | Effect Size: 2.480 | z score | p value: z = 54.71 | p <0.0001 | Number of Comparisons: 188 | Number of Subjects: AD N=12503, CTRL N=8145 |

AD vs CTRL: tau-total (Plasma and Serum) | Effect Size: 1.788 | z score | p value: z = 2.78 | p = 0.00550 | Number of Comparisons: 8 | Number of Subjects: AD N=447, CTRL N=552 |

MCI-AD vs MCI-Stable: Aβ38 (CSF) | Effect Size: 1.127 | z score | p value: z = 2.58 | p = 0.01000 | Number of Comparisons: 2 | Number of Subjects: MCI-AD N=87, MCI-Stable N=144 |

MCI-AD vs MCI-Stable: Aβ40 (CSF) | Effect Size: 1.010 | z score | p value: z = 0.23 | p = 0.81940 | Number of Comparisons: 4 | Number of Subjects: MCI-AD N=187, MCI-Stable N=251 |

MCI-AD vs MCI-Stable: Aβ40 (Plasma) | Effect Size: 1.066 | z score | p value: z = 3.72 | p = 0.00020 | Number of Comparisons: 3 | Number of Subjects: MCI-AD N=308, MCI-Stable N=379 |

MCI-AD vs MCI-Stable: Aβ42 (CSF) | Effect Size: 0.663 | z score | p value: z = 13.95 | p <0.0001 | Number of Comparisons: 19 | Number of Subjects: MCI-AD N=526, MCI-Stable N=881 |

MCI-AD vs MCI-Stable: Aβ42 (Plasma) | Effect Size: 0.807 | z score | p value: z = 0.99 | p = 0.32403 | Number of Comparisons: 3 | Number of Subjects: MCI-AD N=308, MCI-Stable N=379 |

MCI-AD vs MCI-Stable: YKL-40 (CSF) | Effect Size: 1.271 | z score | p value: z = 5.40 | p <0.0001 | Number of Comparisons: 3 | Number of Subjects: MCI-AD N=132, MCI-Stable N=160 |

MCI-AD vs MCI-Stable: neurogranin (CSF) | Effect Size: 1.542 | z score | p value: z = 2.97 | p = 0.00299 | Number of Comparisons: 4 | Number of Subjects: MCI-AD N=190, MCI-Stable N=170 |

MCI-AD vs MCI-Stable: sAPPα (CSF) | Effect Size: 1.091 | z score | p value: z = 1.30 | p = 0.19510 | Number of Comparisons: 3 | Number of Subjects: MCI-AD N=118, MCI-Stable N=169 |

MCI-AD vs MCI-Stable: sAPPβ (CSF) | Effect Size: 1.055 | z score | p value: z = 0.55 | p = 0.58568 | Number of Comparisons: 3 | Number of Subjects: MCI-AD N=118, MCI-Stable N=169 |

MCI-AD vs MCI-Stable: tau-phospho (CSF) | Effect Size: 1.677 | z score | p value: z = 8.05 | p <0.0001 | Number of Comparisons: 13 | Number of Subjects: MCI-AD N=340, MCI-Stable N=645 |

MCI-AD vs MCI-Stable: tau-total (CSF) | Effect Size: 1.730 | z score | p value: z = 14.43 | p <0.0001 | Number of Comparisons: 19 | Number of Subjects: MCI-AD N=481, MCI-Stable N=841 |

# Further Reading

## News

- Serum NfL Detects Preclinical AD, Reflects Clinical Benefit
- Daydreaming Network Serves as Ground Zero for Aβ Deposition
- Among the Healthy, a Positive Amyloid Scan Prompts Questions About Risk
- Can Biomarker Data Predict Individual Risk for Alzheimer’s?